We are searching data for your request:
Upon completion, a link will appear to access the found materials.
Now that we’ve learned about the ecology of living things, let’s return to the two examples we brought up at the beginning of the module.
We first had you imagine that you were an ecologist interested in learning about the natural world and how living things interact with each other as they compete for resources. This type of study would fall under population or community ecology, depending on if you’re studying members of the same species or not. Remember: populations are made up of a single species, and communities are made up of all living things in a specified area.
One example of community ecology is the study of invasive species. One of the many recent proliferations of an invasive species concerns the growth of Asian carp populations (Figure 1). Asian carp were introduced to the United States in the 1970s by fisheries and sewage treatment facilities that used the fish’s excellent filter feeding capabilities to clean their ponds of excess plankton. Some of the fish escaped, however, and by the 1980s they had colonized many waterways of the Mississippi River basin, including the Illinois and Missouri Rivers.
Voracious eaters and rapid reproducers, Asian carp may outcompete native species for food, potentially leading to their extinction. For example, black carp are voracious eaters of native mussels and snails, limiting this food source for native fish species. Silver carp eat plankton that native mussels and snails feed on, reducing this food source by a different alteration of the food web. In some areas of the Mississippi River, Asian carp species have become the most predominant, effectively outcompeting native fishes for habitat. In some parts of the Illinois River, Asian carp constitute 95 percent of the community’s biomass. Although edible, the fish is bony and not a desired food in the United States. Moreover, their presence threatens the native fish and fisheries of the Great Lakes, which are important to local economies and recreational anglers. Asian carp have even injured humans. The fish, frightened by the sound of approaching motorboats, thrust themselves into the air, often landing in the boat or directly hitting the boaters.
The Great Lakes and their prized salmon and lake trout fisheries are also being threatened by these invasive fish. Asian carp have already colonized rivers and canals that lead into Lake Michigan. One infested waterway of particular importance is the Chicago Sanitary and Ship Channel, the major supply waterway linking the Great Lakes to the Mississippi River. To prevent the Asian carp from leaving the canal, a series of electric barriers have been successfully used to discourage their migration; however, the threat is significant enough that several states and Canada have sued to have the Chicago channel permanently cut off from Lake Michigan. Local and national politicians have weighed in on how to solve the problem, but no one knows whether the Asian carp will ultimately be considered a nuisance, like other invasive species such as the water hyacinth and zebra mussel, or whether it will be the destroyer of the largest freshwater fishery of the world.
The issues associated with Asian carp show how population and community ecology, fisheries management, and politics intersect on issues of vital importance to the human food supply and economy. Socio-political issues like this make extensive use of the sciences of population ecology (the study of members of a particular species occupying a particular area known as a habitat) and community ecology (the study of the interaction of all species within a habitat).
We then had you imagine being a future physician seeking to understand the connection between human health and ecology. Humans are a part of the ecological landscape, and human health is one important part of human interaction with our physical and living environment. Lyme disease, for instance, serves as one modern-day example of the connection between our health and the natural world (Figure 2).
More formally known as Lyme borreliosis, Lyme disease is a bacterial infection that can be transmitted to humans when they are bitten by the deer tick (Ixodes scapularis), which is the primary vector for this disease. However, not all deer ticks carry the bacteria that will cause Lyme disease in humans, and I. scapularis can have other hosts besides deer. In fact, it turns out that the probability of infection depends on the type of host upon which the tick develops: a higher proportion of ticks that live on white-footed mice carry the bacterium than do ticks that live on deer. Knowledge about the environments and population densities in which the host species is abundant would help a physician or an epidemiologist better understand how Lyme disease is transmitted and how its incidence could be reduced.
12 Putting It Together
We have shown that the Constitution was a political document, drafted for political purposes by its framers, who worked to replace the “first political system” of the post-revolutionary nation with one characterized by more national power than had existed under the Articles of Confederation. At the Constitutional Convention, the framers reconciled lofty ideals and base self-interests. Through savvy compromises, they resolved cross-cutting divisions and achieved agreement on such difficult issues as slavery and electing the president, though their solution to the former proved to be a temporary one. By getting the Constitution ratified, the framers adroitly outmaneuvered or placated their opponents, in part by promising to pass a Bill of Rights during the first session of the newly created Congress.
The Constitution established a national government distinguished by federalism, separation of powers, checks and balances, bicameralism, and a commitment to government by, for, and of The People. It divided power and created conflicting institutions—between three branches of government, across two chambers of the legislature, and between national and state levels. While the structure it created remains the same, the Constitution has been changed by amendments, interpretation, new practices, and intermediary institutions. Thus the Constitution operates in a system that is democratic far beyond the founders’ expectations. Though it is far from perfect, it is the oldest existing written constitution on the planet, and has provided the basis through which the American people have negotiated both emergent and perennial challenges, from the rise of industrial capitalism, to the practice of democracy itself.
The Constitution always was—and remains—a political document created and developed in political ways for political purposes, and it continues to be the object of political engagement in the twenty-first century.
Microscope lenses are highly complex optical elements that correct for many different optical aberrations. This lecture describes a number of these aberrations and discusses what you should consider when selecting objective lenses as well as other precautions you can take to reduce the effects of aberrations.
- Your confocal microscope has 3 objectives, a 10ࡦ.2na achromate, 20ࡦ.75na fluorite and 40ࡦ.75na apochromat objective. You want to image a sample stained with only a single dye. Which objective is most appropriate and why (note that magnification can be changed electronically in a confocal microscope).
- 10ࡦ.2na achromate
- 20ࡦ.75na fluorite
- 40ࡦ.75na apochromate
- B: 20ࡦ.75na fluorite. Generally, the higher the numerical aperture, the higher the resolution and brightness of your image. Since you work at a single wavelength, you do not care about color correction in the lens and expect a brighter image with a lens that has fewer corrections (in this case, the fluorite).
Pelican’s Perch #19: Putting It All Together
Now that previous columns have littered the runway with theory and attempted to explain manifold pressure, propeller RPM, and mixture, I’d like to present some suggestions for in-flight techniques to extract better performance from these engines. Some of it will be a little repetitive from prior columns in an attempt to make this column more or less “sink-alone.”
First, a Little Housekeeping
Much to my amazement, these columns have been very well received. I’d like to thank you for reading them, and my special thanks to those who have taken the time and trouble to comment in the public message threads, or via email.
Public Responses, Please…
Speaking of which, please note that it is to everyone’s advantage if you leave any questions or comments as public replies, rather than emailing me privately. There are several compelling reasons for this. Most important, everyone learns from a public exchange, and often others will have wanted to post the same comment or question, but didn’t want to appear foolish, or ignorant. Folks, I firmly believe the only dumb question is the unasked one!
Also, a public response and exchange helps keep me straight! More than once, a reader has corrected something I’ve said, or added to it considerably. There are a lot of people out there who know a whole bunch more than I do, some of whom should be writing a column themselves! If I err, or put something badly, and they are spurred into responding, we all benefit.
This feedback is incredibly valuable to me, and to readers who might be misled by my ravings. Finally, these columns and my other activities seem to produce a huge volume of email, and I find myself getting into some fairly deep one-on-one discussions with a number of very nice people, when those discussion would be of much greater benefit in public. I love the email I get, but on the technical issues or issues of broad interest, it would be far more efficient to carry on with them in public, please.
I’m also gratified at the interest in some of the old books I’ve been able to get reprinted, and at the neat responses from those who now have them. So far, there has not been a single complaint or request for a refund, and I’m now into the third reprinting. If any of you should happen to have other old publications that should be made available, let’s talk.
Carbureted Engines, and Fixed-Pitch Props
The single most-asked question I’ve had from these engine management columns has been some wistful variant of, “But what about my 182, which has a carbureted engine?” Or, “What about my 172, with its fixed-pitch prop?”
Folks, I’m really sorry, there’s just not that much that can be done, beyond a few simple tips that have been commonly used for years, like leaning on the ground (see below).
Here’s another tip for carbureted engines only . When operating at full throttle, it is often helpful to back off on the throttle until you see just the tiniest drop in MP (if you have a MP gauge, otherwise look for the slightest drop in RPM), and leave it there instead of fully open. That cocks the throttle plate just enough to set up a slightly turbulent flow, and that helps mix the fuel and air for better combustion. A touch of carburetor heat may help, too.
But I’ve flown a couple of 182s that are simply hopeless, and I’m tempted just to leave the mixture full rich and forget it. Pity, the Skylane is a marvelous airplane, but that TCM O-470 engine has the worst mixture distribution of any powerplant on the planet.
If you’re fortunate to have an all-cylinder engine monitor on a carbureted engine, see if you can operate lean enough to get all cylinders lean-of-peak (LOP) without the engine jumping off its mounts from vibration. I haven’t seen one that would yet, but there may well be some engines that will do it. If so, some of this column may be helpful.
While the knowledge of these things may be helpful in a general way for everyone who flies, it is only the high-performance, fuel-injected engines that allow some of the more sophisticated techniques. Even on those, it almost always takes GAMIjectors to make these tricks work well enough to use them at all. This is the primary reason you have not seen some of this information in the past – it simply hasn’t been very useful, and there’s been no need for it.
Once again, this column will refer entirely to normally-aspirated engines, unless otherwise noted. Yes, yes, I know, there are two or three turbo owners out there who want me to do a column for them, and I’ll get to it, one day, I promise (but I’m not saying when!)
Gentlemen (and Ladies), Start Your Engines
I ran this section by both my editor Mike Busch and GAMI’s George Braly for their comments, and I’d like to quote George’s response here:
“For the record, I think you are a damn fool for doing an article on “starting”! So there! Even GOD doesn’t know how to start some of these engines!”
Now, that was said in George’s usual high good humor, but it’s an excellent comment on the problems in trying to discuss this subject. Behind the scenes, there was a “lively discussion” on this before this column went public. Both Mike and George see a LOT more different engines than I do, and if anything is obvious, it is that there are as many starting techniques and ideas as there are pilots! I have shamelessly incorporated some of their tricks and techniques, too. If they work for you, I’ll take the credit – if they don’t, blame Mike and George.
Starting often seems to be an art form, but there is science involved, too. We know that if we provide air and fuel in the proper proportions and a spark, the engine must run, in-flight. This is also true of the initial engine start, but there are some complicating factors. Let’s take a look at some of them, in no particular order.
For the real techno-freaks, a really important factor is “valve overlap,” (where both the intake and exhaust valves are open at the same time). The pressure (suction) in the manifold is lower than ambient, and lower than that in the exhaust system during starting and low-power operations. Because of this, there is some tendency for the “stuff” in the combustion chamber to get sucked back into the intake manifold. Even some of the exhaust may be subject to this reverse flow during valve overlap.
Another problem is poor vaporization and atomization during the very low cranking RPM. There is a whole host of engine problems that can lead to hard starting, but for our purposes here, we assume a properly adjusted engine.
Cold engines will generally start better with a throttle nearly closed (less air) and relatively more prime than warm engines. To tell the truth, most cold engines start right up without a problem no matter what you do, so I won’t waste much time here discussing them.
Since these “cold starts” require the richer mixture, most aircraft engines are set up for this case, making cold starts the easiest of all. That’s all well and good, but it provides too much fuel for a warm start, which leads to the common problem of “flooding,” which simply means a mixture too rich to burn. Many pilots, when faced with this situation (no start) will think the engine isn’t getting enough fuel, and they will make it worse by priming more, turning the boost pump on, or even pumping the throttle (a bad idea at any time). For this reason, we must preset considerably more throttle with a warm engine to provide more air, and use a lot less prime (perhaps even none) for a proper mix.
If the hot engine is flooded, it is often beneficial to just let it sit for 30 seconds or so. This allows some of the liquid fuel to evaporate into vapor. The chances are that will start the engine instantly on the next try when that vapor gets sucked into the cylinder, and if the fuel pump is pumping fuel and not vapor, a good start will result.
Through experience (or the smell of fuel!), you may recognize a flooded condition. Many pilots will correctly recognize that more air is needed, but they will then make the mistake of quickly shoving the throttle all the way in. This may even work, but quite often, it will cause the mixture to go so quickly from “too rich to burn” to “too lean to burn” that the engine hasn’t a chance to start. So push that throttle in slowly, taking perhaps five seconds for full travel. Chances are that slow opening of the throttle will at some point create just the right mixture conditions, and the engine will fire. That’s probably a good place to leave the throttle for a second or so until the engine starts, but be ready to pull it back quickly to keep from getting a huge surge of RPM.
I absolutely hate hearing an engine start, then instantly run at some very high RPM for many seconds. Some pilots will do that, then bring it back down, but others will compound the error and sit there getting the ATIS, and doing an after-start checklist, with what sounds like 1,800 RPM on a cold engine. Many experts estimate that 90% of all engine wear takes place in that initial cold starting process, when done right . Running at high RPM on bearing surfaces marginally lubricated by cold oil is a terrible thing for your engine, not to mention that the prop blast and noise will make enemies on any ramp.
If you need to prime the engine for starting, use the primer. That’s what it’s designed for. (Well, Duh!) Some pilots will pump the throttle, either before cranking the engine, or while doing so. This is a bad idea, even if it does seem to work sometimes.
Engines with the usual Marvel-Schebler carburetor (as found on Cessna 182s) have an accelerator pump just like your car engine does. Well, let me qualify that, because I haven’t the faintest idea if modern car engines have such devices, and virtually no interest, either! I lost track of what’s what in car engines in about 1964, I think. Dreadful, dirty things, I never did really like working on them. Now airplane engines, that’s another story! Works of art, mostly. But, I digress. Back to pumping the throttle.
Any fuel produced by pumping the throttle is squirted into the carburetor throat, at a point too far from the cylinders to do much good at cranking RPM. Instead of helping the start, the squirt of raw fuel just dribbles down the low point in the intake system and puddles there, creating a major fire hazard in the event of a backfire.
It may also dribble out a drain tube onto the ground, making a puddle of fuel. This becomes an extreme fire risk if any torching occurs. I once very nearly lost a Lambert Monocoupe (an old wood and fabric airplane) from this very thing. Luckily, I saw the smoke, leaned out the door, saw the puddle burning merrily away with flames licking at the belly, and was able to leap out (nearly took the seatbelt with me!) and push the airplane back away from the fire.
Fuel-injected engines don’t have an accelerator pump, so pumping the throttle on one of them does nothing at all but demonstrate ignorance (not stupidity, which is an entirely different phenomenon).
Many problems are mitigated with fuel injection, but others are added. Fuel injection usually puts all fuel right into the intake port (a few, like the superb Wright R-3350, put the fuel directly into the cylinder). It is possible for fuel to run down the “log runner” intakes on flat engines and create some of the above hazards and problems, but it’s far less likely.
Fuel injection is great stuff. But its major problem is the dreaded hot start. If you shut down, then re-start within about 15 minutes, or if you wait an hour, it’ll probably start right up and run just fine. But it is that period of time somewhere in between (30 to 45 minutes after shutdown is generally the worst) that can drive pilots to drink. Nothing seems to work. The problem is that the hot engine “cooks” all the fuel in the engine compartment. It cooks it in the main line coming into the engine-driven fuel pump, the lines after that, and even the tiny stainless lines going from the distribution valve to each cylinder. The result is a system just full of bubbles, and these engines don’t run well on bubbles.
The only effective technique for this is to make sure the mixture control is pulled all the way to idle cutoff (ICO), and then run the boost pump for 30 seconds or so. (On “high” if you have a two-speed pump.) Yes, that’s a long 30 seconds, but it needs to be done.
Note added 06/06/02: Later data strongly suggests the boost pump be run on the highest setting for AT LEAST 60 full seconds, by the clock, and 90 seconds is better.
The electric boost pump is usually located somewhere outside the engine compartment, and thus has no problems with bubbly fuel. What you’re doing here is to circulate cool fuel into and through the engine-driven fuel pump, up to the fuel control unit, and then back through the vapor return lines to the fuel tank. The idea is that when you do crank the engine, the engine-driven pump will really pump liquid fuel, and not starve on bubbles. A normal start should result. While cranking, you might need a quick shot on the boost pump to help the cold fuel get beyond the fuel control unit, and blow out the lines to the fuel distributor (“spider”), and the tiny lines to each cylinder. But make sure it’s really quick , otherwise you risk flooding the engine.
Another hot-start trick is to park downwind (i.e., tail into the wind), and open anything you can on the top of the engine. On Bonanzas, this would be the cowling, on Cessnas, the oil filler cover. The wind will be scooped up into the bottom of the engine by the cowl flaps, and the opening on top allows the hot air out. This helps keep those bubbles from forming in the first place.
Support for this is found in the TCM manuals:
“…During subsequent [after the thirty minute heat soak] starting attempts, the fuel pump will initially be pumping some combination of fuel and fuel vapor. At the same time, the injection nozzle lines will be filled with varying amounts of fuel and vapor. UNTIL THE ENTIRE FUEL SYSTEM BECOMES FILLED WITH LIQUID FUEL, DIFFICULT STARTING AND UNSTABLE ENGINE OPERATION CAN NORMALLY BE EXPECTED.” Page 8-5 TCM Form X30565, August 1990 FAA approved M & O Manual, Model IO-550A, B, C, G.
Some suggest a full-rich mixture, with the boost pump on for a few seconds. That may solve the immediate problem of bubbles in the lines, but it can also pump so much raw fuel through an open intake valve that you’re almost certain to flood the engine, and you might even create “hydrostatic lock” that could damage a cylinder (maybe). Some say that much fuel may also wash the oil off the cylinder walls, which may not be a good thing, and with a non-turning engine, fuel will accumulate, leaving just that much more raw fuel to get rid of. Who knows, I sure don’t, but I’m pretty sure that will create a fire hazard you don’t want. If you absolutely must use this technique, do it while cranking the engine, so at least you have a chance of a start if you happen to get everything “right” during the process.
In any event, for any start with a significant amount of throttle, as soon as the engine fires and the RPM starts rising, move the throttle to the idle position, quickly and smoothly, then increase throttle so as to “meet” the RPM at about 1,000.
Alternator On or Off?
In some circles, there is much ado about whether to start the engine with the G/A (Generator/Alternator) on or off. If you turn it on with the battery before the start, you won’t forget it, but some say that’s too much drain on the battery. Okay, maybe there is a small drain of an amp or two to excite the field, but if your battery is that critical, you probably ought not to fly with it anyway! A similar comment applies to the rotating beacon, which I leave on at all times, so the master switch controls it. (In my opinion, rotating beacons should be on circuit breakers, not switches, so they always turn and flash with the master on.)
A point against leaving the G/A off until after the start, if I may? Watch the ammeter when you do turn it on, and you’ll see an instant huge surge as the G/A instantly tries to recharge the battery after the massive drain of starting. That instant shock is seldom good for shafts and gears, although in theory they’re designed for it. If the G/A switch is on to start with, the system assumes that load much more gently as the starter is released and the engine RPM comes up.
Leaning on the Ground
Finally, you may find it beneficial to lean your engine after start, and for all ground operations. In theory, a properly set up engine will run at “taxi power” without fouling plugs, but the reality is that most general aviation piston engines are set up so rich (for easy cold-starting) they do often foul plugs. You can test this on your own engine by setting minimum idle RPM, then leaning until the engine quits. Watch the tachometer very closely for a small rise just before the engine quits. The more rise you see, the richer your idle mixture setting. On the big radials, this rise should be almost imperceptible, or “barely detectable,” but some of the flat engines call for as much as 100 RPM rise. Again, this is mostly for optimum starting, not running. It’s perfectly safe and often desirable to correct this with manual leaning, once the engine is running.
The downside of leaning on the ground is the very distinct possibility of attempting a takeoff that way, so if you lean on the ground, lean it brutally! You can’t hurt the engine by leaning at “taxi power,” but you sure can cause some heavy damage if you take off with the mixture partially leaned! If you attempt a takeoff while “brutally leaned,” the engine will simply wheeze and die when you try to apply throttle. If you enrich at any time, for any reason, either go right to full rich and leave it there for takeoff, or re-lean it “brutally” once again.
You, there! Get off the brakes!
There seems to be a large group of pilots (including some high-time pros) who set some slightly-high power for taxi, then hold it back by dragging brakes and using differential brakes for steering. That heats up brakes very quickly, destroying some of their effectiveness, wearing them out, and even creating a potential fire hazard. On most of the big airplanes, this is deadly.
Try to control taxi speed with power, without using brakes at all. If you must use some higher power for some reason, allow the speed to build up to “slightly too fast,” then brake it down to “too slow,” and get off the brakes again. That gives the brakes a lot of time to cool in between applications. Brakes like to be applied or released, not “dragged” or “ridden.”
What has that got to do with engine management? Nothing. It’s just one of my pet peeves, and I couldn’t resist sticking it in here!
In the Runup Area
Watch where your tail is pointing!
Yup, you’ve got it, another one of my pet peeves. Consider the wind. Some say this is for cooling, but I’ve seen good data that suggests a downwind runup provides better cooling for a Bonanza! In my opinion, it is far more important to consider where your runup will blow your dust cloud and debris.
If you have properly leaned for ground operation, you might be too lean to get runup RPM, in which case go full rich for runup. When the runup is done, you can leave the mixture full-rich for takeoff, or – if you anticipate a significant delay – re-lean it “brutally.” Remember, we don’t want to risk taking off with the mixture leaned. I don’t care what kind of checklist you use, or how many times “Mixture Rich” is on it, you WILL forget it. This is a perfect example of using a “fail-safe” procedure that will protect you from yourself.
The usual 1,700 RPM for running up most TCM engines (or 2,000 RPM for most Lycomings) is NOT critical. I’ve seen pilots diddle and dawdle trying to get exactly 1,700 but all this does is heat the engine up for no good purpose. Plus or minus a couple hundred RPM won’t hurt a thing, so push it up to “about 1,700” or “about 2,000” and get on with it.
During the mag check, you must see some drop on each mag, but not more than the POH specifies. One exception: If you do a mag check while leaned, the drop will usually be much greater than when the mixture is full rich, perhaps even out of “book limits.” This is expected for runup with a super-lean mixture, and it’s no problem. But, if you’re concerned about that, just do the runup at full-rich mixture! After you runup lean a couple of times, you’ll know what to expect for “normal.” If you have one of the all-cylinder engine monitors, watch the display on that, rather than (or in addition to) the tachometer. That monitor will tell you more about your engine than the simple mag drop.
It is also becoming very clear that the mag check at low power (anything less than cruise power) is not very useful for catching problems it’s nothing more than a quick check to catch major problems like severe plug fouling, a “hot mag,” or a dead plug, or cylinder. This was well-known in the big old radials, where mag checks are almost always performed at about 30″ MP, and up around 2,300 RPM (varies with model).
I would suggest that everyone should get into the habit of checking the mags once each flight, perhaps near the end of the flight, using the engine monitor, at cruise power. Select each mag, and let the engine run there for 30 seconds, watching for EGT/CHT changes and dropouts. This is an excellent check of the entire ignition system, and very often detects problems long before they show up on the usual runup. The engine should run smoothly on each mag during this check, even when LOP.
Cleared for Takeoff
It is my firm belief that all takeoffs in ALL recips should be done at full power, as specified in the POH. I cannot think of a single exception.
(Pop-quiz question: At full takeoff power, would you expect higher CHT on a cold day, or a hot day? See answer below.)
If the book says full throttle and 2,700 RPM (or 44″ and 2,800, or whatever) then that’s what you should use. Get the governor setting tweaked so the prop delivers the book RPM on takeoff, as measured by an accurate tachometer. I won’t settle for more than 50 RPM either way, and it can be held to even tighter limits, with a little effort.
Similarly, get the fuel flow tweaked so that you get ALL the fuel flow the book allows. This is important for cooling . If it’s a hair over, that’s fine, but don’t settle for less than “book” fuel flow.
Some are confused by the fact that some engines require the use of a boost pump for takeoff, and some do not. The big radials have carburetors that are set up to have auxiliary boost pumps on for takeoff, and most of them require them to be switched on to cover the possible failure of the engine-driven pump. On the other hand, many general aviation flat engines are set up to not use the auxiliary boost pump for takeoff, while others have a two-speed pump that should be set to “LOW” for takeoff. Simple solution, just “Do what the book says!”
On flat engines, “LOW” boost may be helpful on hot days, but if you use it, you may need to manually lean the mixture to prevent power loss. “HIGH” boost is generally prohibited on flat engines, because it will deliver enough fuel to flood the engine out at high power. If the engine-driven fuel pump does fail, then it’s time to use “HIGH” boost.
On a climb with hot fuel in the tanks, and a hot day, you may need “LOW” boost to keep bubbles from forming, and making the engine surge and sputter. Go ahead and turn it on, but you may also need to lean just a bit more to keep the fuel flowing at the proper rate.
Generally speaking, flat engines with a single-speed auxiliary boost pump with only an “ON” and “OFF” setting are really in the “HIGH” position when on, and for this reason, “ON” should not be used unless clearly needed for a fuel pump failure.
On normally-aspirated engines (carbureted or fuel-injected) when operated at high altitudes, it is necessary to lean for takeoff, because the enrichment function will cause a “too-rich” mixture. The TCM IO-550 is sometimes an exception to this, as some models have an “altitude compensating” device. In my experience they don’t work very well, and I have mine tweaked to not do very much, leaving mixture control to me.
Just how to do that leaning for takeoff will produce heated debate in any airport lounge. But it’s not that hard! Look at the TCM power curves in “Mixture Magic” and notice that the BHP curve is pretty flat in the high-power range. That tells you that the mixture setting isn’t critical – we just need to get it in the ballpark. All we need to do is get rid of some of the “extra enrichment” and we’ll be somewhere on the flat part of that curve, and that’s about as good as you can do. Personally, I just go full forward with everything, then on the roll, I grab the mixture control and make a gross movement “too far,” feeling for the power loss, then I shove it back in to the point where it “feels good,” and let it go at that. We do not need to be super-precise here! Isn’t that simple? The momentary “too lean” mixture won’t hurt a thing, especially at the high elevation when the engine is putting out a lot less than full rated power.
To pound on the poor dead horse one more time, y ou are not “saving” an engine by using less than full power for takeoff, you are probably hurting it! All high-performance aircraft engines have some means to greatly enrich the mixture at takeoff power. When you attempt a takeoff at partial power, you often defeat this, and both EGT and CHT will be higher, often much higher before the gear comes up. Full power will also get you higher, faster, sooner, and this is good from a general safety standpoint. Also, CHT rises continuously throughout the takeoff in all engines, because at low speed, there is insufficient airflow for cooling. The quicker you can get to an airspeed that does provide good cooling, the sooner that CHT will stop rising, stabilize, and even start down again. The operators of one large warbird persist in taking off with reduced power, and incur the double whammy of slow airspeed acceleration on the runway and a less-than-optimum mixture, leaving them looking at redline CHTs by the time the gear is coming up. They “just don’t get it.” If they’d just go ahead and use the bloomin’ power the manufacturer specified, CHT would probably not even be an issue!
Partial power takeoffs in a multiengine aircraft are really dumb, because performance is greatly reduced, and the power setting greatly complicates the engine failure procedures. You may think you’re good enough to add power on the good engine while simultaneously handling the failure of the bad one, but trust me, you’re not. You’ll blow it every time. Again, you do not hurt an engine by operating it at full takeoff power as specified by the factory! This will generally keep the engine cooler .
The only case I know of where reduced power (thrust) is safe and useful is in jet aircraft, where the CHT problems do not apply. The fuel control units are so good that proper mixtures are maintained at all settings, and the aircraft have such an excess of thrust that the engine failure case is near-trivial. In fact, when briefing a “reduced thrust” takeoff, I make it very clear that if we do lose an engine, we will normally NOT advance the thrust on the others at all. If that doesn’t seem safe, we go to the next higher available thrust setting.
Oh yes, that takeoff CHT question above? CHT will go higher on a cold day. This is because the normally-aspirated engine produces more power with cold air (denser air, more fuel flow), and even more importantly, the engine is producing that slightly higher power at a leaner mixture (more air, same fuel). More power and a leaner mixture equates to higher CHT. This normally overrides the small effect of colder cooling air.
Climbing to Altitude
My pet peeve in climbs is using too low an airspeed. One famous training organization insists on using 95 knots to 1,000 feet AGL in a Bonanza, most of the twin operators push for a climb at or near the blue line, and virtually all the transport operators use a V2+10 climb with all engines operating. I think it’s stupid to climb that slowly in any airplane (including jets, just so I offend everyone here!) once actual obstacles are cleared .
95 knots in a Bonanza runs the engine temperatures up. Yes, they may remain within limits, but why get any hotter than necessary? Additionally, the nose will be so high you simply cannot see where you’re going – you’re blind to traffic that may very well be in your path. Finally, I firmly believe that if a total engine failure occurs at 95 knots in a Bonanza below a few hundred feet, most pilots will stall before they can get the nose down enough to maintain flying speed. If they just happen to succeed at that, they will probably end up so slow and descending so steeply that there isn’t enough energy left to flare. A crash, and an ugly one, is all but inevitable. At such low speeds, with the usual nose-high attitude, a recovery from an engine failure is very nearly an acrobatic maneuver, and not one pilot in a thousand has practiced it with any realism. It’s a bad deal .
For this reason, I prefer to see a very early shift to a higher climb speed once real obstacles are cleared. In the absence of real obstacles , I set up a gradual climb right from liftoff, pulling the gear up as soon as I am well clear of the ground (Oh, boy, I’m gonna get mail on that one! Hmm, might be a good subject for a column?) In the Bonanza, I accelerate to about 120 knots, reaching that speed by the time I’m at 100 feet AGL, or 200 AGL. This gives much better cooling, much better visibility, and makes the engine-out case far more manageable (single or twin). Yes, yes, I know, I’ll be at a slightly lower altitude when the engine quits, but not as low as you might expect. Some are very fond of quoting the sharp rise in drag with higher speed, but an often-missed factor is the improvement in prop efficiency that also takes place. The real result is that the actual climb rate on a Bonanza will suffer very little. The angle of climb (gradient) drops, of course, and it is possible for this to become a terrain clearance issue.
I’d rather have just a little extra speed here, and I accept the slightly lower altitude.
Power reductions after takeoff have many considerations and variations, so it’s a little hard to generalize. In all cases, use full takeoff power until the “flight situation” has “stabilized.” This is a very subjective point, since some pilots won’t be fully stabilized until the airplane is in the hangar and they’ve had a transfusion of some type of spirits. In general, though, the gear should probably be up, the airspeed stabilized, some altitude beneath the wings, and the workload of takeoff should have abated somewhat. (On a low-IFR departure, perhaps you’ve turned to your initial departure heading and broken out on top of the low stratus deck.) That’s the time to set whatever lesser power the limitations section of the POH calls for. Some of the larger engines have a “METO” (Maximum Except TakeOff) power, which must be set within one, two, or even five minutes of takeoff. Some go further, with an even lower setting for “Climb.”
Many of our flat engines have no such limits, and are rated for full takeoff power “essentially forever.” Some have RPM limits, and those should be observed. With normally-aspirated flat engines, the increasing altitude we normally see automatically reduces the power within a couple of minutes, so it is almost silly to retard the throttle, then keep adding it back within minutes.
There is a fast-growing major problem, however, and that is noise. A primary reason airports are closing by the dozen is noise at takeoff power, or more correctly, noise at takeoff RPM. It is irritating , sometimes even to people who love airplanes! In Seattle, where I live, there is a constant procession of floatplanes overhead at 1,500′ to 3,000′, and many of these pilots and operators don’t have this picture, because they’re boring holes in the sky on sightseeing and training flights at 2,500 RPM or more, engines just screaming. I can only imagine the effect on people who don’t like airplanes at all, or those who feel endangered by them.
For this reason, I suggest pulling off a couple hundred RPM as soon as possible after liftoff, no matter what the book says. The difference in noise from my IO-550 and its three-bladed McCauley prop is dramatic, at the cost of just 15 HP. The Germans are so anti-noise they made Beech/TCM deliver a “special” engine on Bonanzas to be operated there, and the only change is a limit of 2,500 RPM (instead of 2,700) on the same engine. Since Beech had to do certification testing for this, we know what the HP is. The IO-550 is also quite happy at 2,500 RPM for a lot of reasons, so it’s a good trade-off, I think. I usually do that as soon as the gear is up.
At the risk of being accused of beating that poor dead horse again, please , please do NOT reduce the MP to 25″ after takeoff or for climb, as has so long been the accepted practice on the flat engines! EGT and CHT will go up, not down. You are not doing your engine any favors, and you may be hurting it.
Unless there is some limitation in the POH, climb at WOT (Wide Open Throttle), 100 or 200 RPM below the max (for noise), full-rich mixture and cowl flaps open (cooling), or cowl flaps as called for in the manual. You can’t go wrong doing this. Monitor your CHTs, preferably on a modern all-cylinder monitor, and if any CHT goes above 380°F to 400°F, do something about it . There are several things you can do. The most effective one is to lower the nose and increase airspeed. Open the cowl flaps if they’re not already open. Enrich the mixture, if you’re ROP (or lean it more if LOP). If you’re already full rich, this might be a good time to turn on the boost pump to “LOW” (if available) or “HIGH” to increase the fuel flow. With the boost running, you may get too much fuel, and the cure for this is to lean a little. All of these will reduce climb performance, but unless you’re about to hit something, it always makes sense to take the “hit” on climb rate in order to keep the engine nice and cool.
No matter what your POH says, limitation or otherwise, I suggest you consider 400°F an absolute redline CHT on any cylinder with a bayonet probe feeding an all-cylinder monitor (spark plug thermocouple types may show higher). There is mounting evidence that factory limits on CHT are much too high. In any event, try to keep the time above such temperatures at a minimum.
Leaning for Climb?
The question is not whether to lean for climb, but when to start the leaning. Many POHs will state pretty strongly that no leaning at any power setting be done below some fixed altitude, often 5,000 feet, and then in “cruise” only. This is patently absurd, but some people carry that to ridiculous lengths, even saying that it applies on the ground, for taxi operations!
I suspect stuff like this comes from a quick meeting with a non-flying lawyer present, who insists that the POH must be written for the lowest common denominator among pilots, the utter moron. Various suggestions get kicked around, and finally someone observes, “Well, you won’t hurt the engine, or reduce safety with full rich below 5,000 feet, and you won’t hurt the engine, or reduce safety, by leaning above 5,000 feet.”
“Aha! That’s easy, we’ll just put in one line saying ‘Lean only in cruising flight above 5,000 feet.'” Simple, effective, undeniably safe – who can argue? The whole mixture discussion is cut to one simple, clear line, over which the manufacturer will never get sued.
Even GAMI is conservative about this subject, as misuse of the mixture control can indeed ruin an expensive engine, and I suppose they could be sued. Be forewarned, my defense in court will be, “You did what ? You took advice from some nut on the Internet? Stupid!”
Can I reduce “leaning in the climb” to one simple sentence? Sure, it’s easy! Like the book says, “Lean only in cruising flight above 5,000 feet.”
But even this can get you in trouble! Take the case where you’re climbing out, ROP or even full rich. Hot day, hot fuel, and the engine-driven fuel pump is hot enough to heat the fuel even more. Bubbles begin to form, and these move through the lines, to the combustion chambers. The fuel flow sensor won’t detect the bubbles, so you show the same old fuel flow, but with all the air bubbles, it’s actually a very lean mixture. This can drive one or more CHTs up very quickly. If you have a cylinder that runs hot already (as most #2 jugs do on Bonanzas), this may be just enough to drive that jug into detonation.
On the other hand, if you are running LOP, this will drive the CHTs down, and cause a power loss. Nice to have the settable warnings on an engine monitor, as the JPI has.
To operate lean-of-peak (LOP) during climbs, or in cruise below 5,000 feet, you need at least all of the following:
- A thorough knowledge of just what is happening when you lean. How thorough? Well if anything in “Mixture Magic” is a complete mystery to you then don’t do it!
- A good, accurate all-cylinder monitor showing at least EGT and CHT in a digital display (preferably with one-degree resolution),
- A fuel-injected engine with GAMIjectors installed,
- A higher climb speed than you may be used to.
If you lean aggressively during climbs at the lower altitudes without all those factors present, you’re playing with fire. It can probably be done, but I’m not sure you can learn how, safely.
When I’m ready to go LOP, I’ll have 120 knots, and I’ll start out with WOT (Wide Open Throttle), 2,500 RPM, and the mixture in full rich. I’ll have a “stable situation” with a low workload. (I don’t fool with things like this during an instrument departure, for example.)
(If you are irritated by TLAs (Three-Letter Acronyms) I’m afraid you’ll have to deal with it. The “WOT” and “LOP” terms are spreading rapidly, and the GAMI folks are even tossing around “WOTLOPSOP” for “Wide Open Throttle Lean of Peak Standard Operating Procedure.” But, I digress.)
I’ll set my engine monitor to show the hottest CHT. Now if you’re going to lean at this point, you can’t pussyfoot around and do it slowly! If you do, the EGT will soar, and the CHT will follow fairly quickly, as you move the mixture control. You must grab it, take a deep breath (the first time), and firmly do the BMP (Big Mixture Pull) moving the fuel flow fairly quickly to LOP in about one or two seconds. In my engine, pulling it right back to 16 GPH will be very close, giving me about 1,450°F EGT, and about 370°F CHT. How can you tell how far to pull it that first time in yours? A major immediate indication will be a noticeable loss of power. The next indication will be the trend of that CHT. If it rises, you need to lean more. If it falls quickly, you leaned too much. If that sounds backwards, remember you’re now on the lean side of the mixture curve, where “leaner is cooler, richer is hotter.” This concept may take considerable “retraining” because you’ve probably spent your whole flying life thinking just the opposite. Even the FAA exams teach that “leaner is hotter” but that’s true only on the rich side of peak.
Okay, let’s assume you’ve successfully transitioned to LOP. The idea now is to set the mixture so that the CHTs ride somewhere between 350°F and 380°F. I think the upper limit is important (see “Mixture Magic”). The lower limit is somewhat arbitrary, but in my opinion, you lose too much power if you lean it below that. By using a lower number at first, you’ll have a chance to look carefully at all your CHTs, and get a feel for how the engine likes this mode of operation. Personally, I shoot for about 380°F in the climb.
CHT reacts very slowly, however. Most people find it easier to use the EGT for changes, as it reacts instantly. Then see how that affects the CHT. With just a couple tries, you should be able to determine an EGT that will give you pretty close to the target CHT thereafter, you can just set the EGT. I use 1,450°F on the lean side for climb, yours may vary.
It is very important to control your airspeed! Pick a climb airspeed and stick to it! If you let the IAS drop off, your CHTs will certainly rise.
Some airplanes will go all the way to cruise altitude with no further adjustment. With some, the CHT may rise, or it may fall. You may need to make one or two adjustments during the climb, remembering that on the lean side, “leaner is cooler.” Tweak the EGT by 20°F, then later see what that did to the CHT.
Once you reach cruise altitude, very little adjustment is needed. Pick the RPM you want (I use 2,500 at altitude, 2,100 when very low) and let the airspeed build. As it builds, you’ll see the CHTs drop from the increased cooling. If you like, “toss another log on the fire,” and enrich the mixture to bring the CHT (and the power) back up to the level you want, but always less than about 380°F.
As far as the engine is concerned, we could skip this “cruise” section entirely. By just leaving everything alone (or just tweaking the CHT) when reaching cruise altitude, you’re all set in the “LOP go fast” mode. But there are other considerations. You may not want the “go fast” mode – you may need more range, for example.
In my opinion, it is utterly stupid to set some arbitrary percentage of rated power for cruise. That’s getting the cart before the horse. What you need to know is the airspeed you need for the purpose of this flight , then set the power to produce that airspeed .
This is not a new concept. Charles Lindbergh used it in 1927 in the “Spirit of St. Louis,” and later taught the technique to military pilots in WWII. Airlines and military transports used it as a normal procedure in prop aircraft for decades, and modern jet transports do it, too. In cruise flight in jets, we haven’t the faintest idea of how much “power” or “thrust” we’re pulling, all we’re concerned with is the SPEED we want, with the only real engine limit being – would you believe – EGT?
Try this yourself, with your airplane at gross. Make a table with four columns, and label them “Fuel Flow,” “IAS” (Indicated Airspeed), “TAS” (True Airspeed) and “MPG” (Miles Per Gallon”). Next time you’re flying pretty close to max gross for your airplane and the air is smooth, set up “LOP Go Fast” mode as above, and note the fuel flow and IAS. (Do the calcs for the “TAS” and “MPH” later.)
(Yes, you need a good fuel flow indicator for this.)
Next, reduce the fuel flow a few tenths of a gallon, and let things stabilize for a minute or so. Note fuel flow and IAS again. Repeat this at intervals of a few tenths of a gallon. At some point, the engine will become slightly rough, but don’t panic, just note the fuel flow. Reduce the RPM by 200 or 300 (which will reduce the fuel flow even more,) then set the fuel flow back to the setting you noted, and resume gathering your data. At some point, the engine may get rough again, so run the RPM down another couple hundred.
What you want is a table that shows LOP fuel flow and the corresponding indicated airspeeds. Because horsepower varies almost linearly with fuel flow when operating LOP , it doesn’t really matter what the RPM is for this exercise – the results are “close enough.”
When time permits, run the numbers to convert IAS to TAS for each line, then simply divide the TAS by the fuel flow for each line, and you’ll have “MPG.”
If you really want to do it right, do this at low altitude, again at some intermediate altitude, and at the highest normal altitude you use. For me, that’s 1,500 feet, 9,000 feet and 19,000 feet. Plot this data out on graph paper, with fuel flow across the bottom, and TAS up the side, or better yet, let Excel’s superb graphing feature do it for you.
You’ll probably find that by decreasing your IAS 10 or 15 knots, you can get a pretty fair improvement in MPG, and this is worth knowing. Your best mileage will be at a very slow speed, but very few of us have the patience to do that. You must also take wind into account – you’ll need more IAS/TAS in a headwind, and less in a tailwind.
(I apologize for not having the data and a nice neat color chart showing all this. I had one, but my computer had other ideas about it, and lunched it.)
How do we use all this? For the normal flight, I just set “LOP Go Fast” and forget it. But when I see that fuel might be critical, and I want to make it nonstop, I’ll reduce the power (and thus IAS and TAS), and see how that looks for a time. If I know I’ve got a headwind, forget it and set “LOP Go Fast,” since it’s not worth the effort. With a little knowledge of the data, you can even crank it into your flight planning software, and play some “what if” games.
Descents and the OWT of “Shock Cooling”
I know of no real data, old or new, that supports any of the theories about “shock cooling” being particularly harmful. I think it’s a load of poppycock, invented to give pilots something to sound knowledgeable about, when talking to the less experienced, and it gives CFIs yet another procedure with which to hammer on trainees.
On the contrary, there seems to me to be considerable evidence that “shock cooling” is not particularly harmful. Airplanes flown regularly (and hard) seem to be the ones that regularly go to TBO and beyond. Some of these are flown in the harshest conditions found on earth, such as the arctic, the jungles and the desert. Above all, training aircraft are subjected to the very operations that cause the most severe shock cooling on virtually every flight, with constant simulated engine failures, aborted takeoffs, long power-off glides and sudden applications of full power. Aircraft that tow gliders routinely go to TBO, and they are doing constant full-power climbs at low airspeeds, followed by steep, power off descents, often at high speed. How about acrobatic aircraft, which go from wide open to power off, in all attitudes, at all airspeeds, show after show, sometimes multiple shows per day? Bob Hoover goes from a full-power setting into instant feather at very high speeds (probably beyond redline), flies for several minutes that way doing his wonderful act, then fires them up and within seconds, goes to full power again. He reports that he routinely goes to TBO!
Even if there is something to the various “shock cooling” theories, I don’t think it’s the major factor many make it out to be. On the other hand, if we are to worry so much about shock cooling, what about “shock heating”? Every takeoff involves going from near-idle power to full power within a few seconds.
For a normal descent from any altitude, I simply start down without changing any of the engine controls. The increased speed will cause the CHTs to drop very slowly. If you wish, you could enrich (from LOP), or lean (from ROP) to modulate this, but it hardly seems worth the effort, to me.
At some point during the descent, this will produce an IAS that is “too fast,” either because of turbulence, or the yellow arc on your airspeed indicator. Leaving the mixture alone, bring the RPM back a couple hundred, and watch, maybe tweaking the mixture as needed to keep the temps up, but probably not. When the RPM is back “too low” or the prop hits the high-pitch stops and won’t hold the RPM down, and you’re going too fast, that’s the time to start back with the MP. That first reduction from WOT will move it out of the “power enrichment” range, so you may notice a sharp drop in EGT and CHT (if LOP), or a rise (if ROP.) You could control this with mixture, but I find it’s usually easier just to pull five to ten inches off the MP and be done with it. That gets the engine down into a low power range where nothing will hurt it.
Some pilots spend a lot of money installing speed brakes, because they want the ability to get down faster, without pulling much power off. If I’m cruising at high altitude, and ATC wants me down, or I need to dump it through a layer of potential icing conditions, I’ll put the gear down (with the 156 IAS Vlo on my Bonanza, the gear is a wonderful speed brake), pull off five or ten inches of MP, and head on down at 2,000 FPM or better. Does my engine get cold? No, because I’ll enrich to peak EGT, then another 50 to ROP, and that will keep my CHTs nice and toasty. They may slowly drop a bit, but I’m not concerned about that. In other words, at lower power settings, I’m running at the “best power” point on the mixture curve for temperature control . This gives a surprising amount of temperature control, too!
In the Pattern
I’m a firm believer in doing as little as possible in the pattern, basically gear and flaps only. Everything else should be done long before you arrive in the pattern, maybe 10 miles out, or more, and any and all checklists (written or mental) completed. The traffic pattern deserves your full, undivided attention outside the cockpit .
The classic pre-landing procedure method calls for the prop at a very high RPM setting, and the mixture full rich for a potential go-around. But that high RPM is totally unacceptable around most of today’s airports (see previous comments on noise.) I strongly suggest you keep the RPM at some very low setting (1,800 to 2,100 on mine, please). When you are on very short final, with the speed decreasing, and the power coming back, that’s a good time to push the prop control up, because the prop has probably dropped out of governing range, anyway. Even better, leave it alone.
A somewhat more controversial technique is to leave the mixture leaned where it was in cruise or descent throughout the approach and landing, and all the way to the hangar. I like to do this, because it avoids dumping a whole bunch of cold fuel into the engine when it’s not needed, and it leaves me pretty well set for the taxi to the hangar. But it does have its risks. The one thing that must be done with this technique is to train, train, train for the go-around with MIXTURE, PROP and THROTTLE. This is not a bad training idea no matter what technique is used.
I use the GUMP check myself (Gas, Undercarriage, Mixture, Prop), and when I get to the last two I say (usually aloud) “Mixture is leaned, prop is 1,800, a go-around will require MIXTURE, PROP and THROTTLE.”
Remember, if you get in trouble doing any of this, you’ll have to tell the judge that you were following the advice of some nut on the Internet.
Use Your Head
The techniques in this column are not for unthinking pilots who prefer the easiest way of doing everything, and who can’t be bothered with advanced techniques. If you’re not willing to study and learn WHY these procedures do what they do, then don’t use them ! If you cannot use them regularly , in a careful, considering manner, then don’t use them !
On the other hand, if you are prepared to study and understand how that incredible piece of machinery under your cowling really works, to use judgement about when these procedures might be appropriate, and if you’re willing to put up with the abuse of those who prefer dogma over logic, this column is for you.
Enjoyed all your articles on MAP, RPM and mixture. As you noted, current cockpits with EDMs measure LOP and ROP indirectly via EGT & CHTs. The best way to set air/fuel mixtures directly is thru a Lamda or Air/Fuel gauge which use a oxygen probe in the exhaust stack as have used on cars with EFI/ECUs for more than two decades. Unfortunately, the lead in 100LL quickly contaminates the oxygen sensor and another reason to get lead out of aviation fuels so this technology can be adapted to aircraft engines. I have the advantage of flying a experimental Super Legend with a low compression Continental Titan x-340 and can burn lead-free 93 octane MOGAS. So, minus the lead, can adapt a dual Lamda gauge with dual oxygen sensors, to monitor each of the twin custom tuned exhaust stacks. For take off and climbs use a best-power setting of 12.7:1. In cruise, make the rapid mixture pull to the lean-of-peak setting of 16:1. For peak output for best speed use 14.7/1. I still use my EDM for all the other information it provides for monitoring and trouble shooting but the Lamda gauge sure cuts down the tweaking of the mixture knob. The Bosch oxygen sensors are more resilient to lead contamination and as long as you keep the engine LOP this contamination is minimized. Addition of 4 oz of Marvel Mystery Oil to 10 gallons of 100LL helps reduce the contamination for longer sensor life when forced to buy 100LL on cross countries.
Explaining the weight plateau
The overlapping physiological changes that occur with weight loss help explain the near-ubiquitous weight loss time course: early rapid weight loss that stalls after several months, followed by progressive weight regain 32 . Different interventions result in varying degrees of weight loss and regain, but the overall time courses are similar. As people progressively lose more and more weight, they fight an increasing battle against the biological responses that oppose further weight loss.
Appetite changes likely play a more important role than slowing metabolism in explaining the weight loss plateau since the feedback circuit controlling long-term calorie intake has greater overall strength than the feedback circuit controlling calorie expenditure. Specifically, it has been estimated that for each kilogram of lost weight, calorie expenditure decreases by about 20 kcal/d whereas appetite increases by about 100 kcal/d above the baseline level prior to weight loss 31 . Despite these predictable physiologic phenomena, the typical response of the patient is to blame themselves as lazy or lacking in willpower, sentiments that are often reinforced by healthcare providers, as in the example of Robert, above.
Using a validated mathematical model of human energy balance dynamics 27,31 , Figure 2 illustrates the energy balance dynamics underlying the weight loss time courses of two example 90 kg women who either regain (blue curves) or maintain (orange curves) much of their lost weight after reaching a plateau within the first year of a diet intervention. In both women, large decreases in calorie intake at the start of the intervention result in rapid loss of weight and body fat leading to a modest decrease in calorie expenditure that contributes to slowing weight loss. However, the exponential rise in calorie intake from its initially reduced value is the primary factor that halts weight loss within the first year. In contrast to the modest drop in calorie expenditure of less than 200 kcal/d at the weight plateau, appetite has risen by 400 kcal/d and energy intake has increased by 600 kcal/d since the start of the intervention.
Mathematical model simulations of body weight, fat mass, energy intake, energy expenditure, appetite, and effort for two hypothetical women participating in a weight loss program. The curves in blue depict the typical weight loss, plateau and regain trajectory whereas the orange curves show successful weight loss maintenance.
These mathematical model results contrast with patients’ reports of eating approximately the same diet after the weight plateau that was previously successful during the initial phases of weight loss 33 . While self-reported diet measurements are notoriously inaccurate and imprecise 34 , it may be possible to reconcile such data with objectively quantified increases in calorie intake. It is entirely possible that patients truly believe they are sticking with their diet despite not losing any more weight or even regaining weight.
The patient’s perception of ongoing diet maintenance despite no further weight loss may arise because the physiological regulation of appetite occurs in brain regions that operate below the patient’s conscious awareness 37 . Thus, signals to the brain that increase appetite with weight loss could introduce subconscious biases such as portion sizes creeping upwards over time. Such a slow drift upwards in energy intake would be difficult to detect given the large 20% fluctuations in energy intake from day to day 38,39 . Furthermore, a relatively persistent effort is required to avoid overeating to match the increased appetite that grows in proportion to the weight lost 31 . For example, the model-calculated intervention effort for the simulated patient who experiences the weight plateau at six months followed by weight regain ( Figure 2 , blue curves) maintains more than
70% of their initial intervention effort until the plateau. Perhaps self-reported diet maintenance before and after the weight plateau is more representative of the patients’ relatively persistent effort to avoid overeating in response to their increased appetite 31 . New technologies using repeated weight monitoring can be used calculate changes in calorie intake and effort over time 40 and help guide individuals participating in a weight loss intervention 41 .
THE MACHINE LEARNING APPROACH
Setting up a machine learning model such that the predictions are valid and accurate can be a daunting task, not substantially different from developing a biostatistical regression model. Below, we describe the end-to-end process of machine learning (Figure 1), with examples in epidemiology and public health.
The literature includes numerous approaches for identifying appropriate sample sizes for machine learning models (6, 21, 47). However, sample size estimates are difficult to compute because machine learning models are largely algorithmic based. Most do not utilize frequentist statistical measures such as p-values, nor do they focus on effect sizes, two concepts central to the traditional calculation of sample sizes.
For unsupervised models, sample sizes should be based on the research question and inherent variability of the data under consideration. Because unsupervised models can be used for hypothesis generation or data reduction, sample size calculations may be unnecessary. For example, very small data sets (13 unique individuals) have been used to detect clusters of patients with similar inflammatory markers among hospitalized patients with pneumonia (70).
In short, many loose recommendations and heuristics are available for machine learning sample size estimation. Publications in machine learning sample size estimations are often specific to a discipline. For example, genetic epidemiology may require smaller data sets (e.g., <100 rows) (12) as compared with a moderate size (several hundred cases) necessary for a behavioral/cognition outcome evaluating functional magnetic resonance imaging (fMRI) (11). However, it is important to note that the optimal sample size is dependent on the data available and is based on the number of rows and the number and quality of features. If features included are redundant or not predictive of the label, the model may be inaccurate regardless of the volume of features. Furthermore, if there are many features but few instances of the label, then models may have difficulty matching patterns to the label for the full feature space and the model will be unlikely to function appropriately in production (29). Much like any modeling scheme, results can be generated regardless of the sample size. In machine learning, for the results to be accurate and generalizable, the overall sample size needed should be carefully considered a priori and may be much larger than anticipated (65).
Parsimony is a central tenet of regression model building in epidemiology to prevent overfitting. Selecting the relevant predictors with the appropriate level of explanation is critical to the model's success. Like regression modeling, overfitting is a major concern in machine learning modeling parsimony is accomplished through feature selection in which the features in data sets are thoughtfully chosen for the model. This approach is especially important for machine learning models because they are applied to data sets that were collected for reasons other than a specific hypothesis, as is the case with electronic medical record data (32) or genetic epidemiology data (40). These types of data sets typically contain a large number of features compared to regression models many of them are irrelevant to the model being constructed. One example outlines a technique to detect and limit the set of variables to be used for modeling (known as the feature set) in antigen discovery for vaccinology (18), which could be applied to any ‘omics data set.
The difficulty in the selection of a parsimonious feature set is more complex than just the feature set's impact on a particular study outcome (9). There are many other reasons to reduce the number of features in one's training data set prior to using a machine learning model. First, the model will train faster, which is particularly attractive with complex modeling schemes under local computation as opposed to cluster computing. Second, reducing the number of redundant features or features that do not affect the outcome may decrease the likelihood of overfitting the model.
Feature selection can be done in numerous ways including selecting clinically meaningful features, simple correlations between features, and feature importance scores. Other machine learning models such as least absolute shrinkage and selection operator (LASSO) regression may also be useful for feature selection (62, 69). Genetic algorithms for feature selection have become popular and have been used for various purposes (63), including to understand the impact of uncontrolled comorbidities on clinical outcomes in hospitalized patients with pneumonia (3). Regardless of the method used for feature selection, investigators have suggested that the accuracy and stability of the model should be considered when using feature selection algorithms (21). Otherwise, these models risk overfitting.
In the era of ‘omics data, the feature set provided to epidemiologists for analytics has expanded substantially (22). Many approaches for selecting features have been proposed (24, 25). One example is ranked guided iterative feature elimination (RGIFE), which shows promise for identifying enhanced clinically relevant biomarkers (26) (see the sidebar titled Ranked Guided Iterative Feature Elimination for Feature Selection). Regardless, much like in explanatory regression model building, strict automation of feature selection is likely not an appropriate solution on its own. In nearly all areas, domain experts should be enlisted to assist in feature selection for meaningful models to be developed.
Issue: Data sets with a large number of features are difficult to use because relevant features are difficult to identify.
Solution: Ranked guided iterative feature elimination (RGIFE), an algorithm that uses cross-validation to identify relevant features in classification scenarios, is proposed. RGIFE first estimates the performance of a model with the original feature set with k-fold cross-validation. The model then ranks the importance of the features to the classification task. From there the model removes attributes from the end of the feature rank (lowest ranking features) and runs the model again. Reduced feature sets that perform within a tolerable level are accepted until the performance of the model fails below a specified threshold. The performance of this feature selection model was compared with several commonly used feature selection methods.
Conclusion: RGIFE provided similar prediction performance with few features for several cancer-related data sets.
Creating or engineering new features from available data to capture latent effects is another important facet of machine learning. Historically, feature engineering has been a manual and laborious process, limited by many factors including the mathematical expertise, time available for analysis, and domain knowledge of the study team.
Simple feature engineering, such as taking the logarithm of a continuous variable to change its distribution or aggregating two variables to account for multicollinearity, is sometimes necessary in traditional regression modeling. For example, the performance accuracy of machine learning models to predict early sepsis was improved by multiplying a shock index by age to derive a new feature for regression (19).
Feature engineering is becoming more complex, with the potential to uncover latent effects that would not be accounted for otherwise. One novel automated feature engineering approach is deep feature synthesis, which combines multiple feature transformations and aggregations, of any type or complexity, to create new features (33). Individual variables are precursors to these new deep features, created through primitives, mathematical formulas used to transform or aggregate. Primitives can range from simple to complex, including means, sums, principal components, or even predicted probabilities or error terms from traditional regression models. Aggregations are very useful for longitudinal features, whereas transformations are typically used for time-invariant features.
In our experience, approaches to feature engineering often merge with feature selection, especially for longitudinal data sets. For instance, in clinical epidemiology, when selecting features, investigators often must limit which laboratory values to include in a model because the number of time points and frequency at which laboratory data are collected during a hospital stay vary.
External data sources are becoming important components of accurate machine learning models. Affixing data collected outside the primary data set provides machine learning algorithms additional features from which to learn. In fact, researchers have suggested that different machine learning algorithms are unlikely to provide a substantial improvement in model performance if the same feature set is used for each (19, 25).
External data sets can be combined with primary data sets, including, among others, geographic location, weather data, and aggregated population statistics. For example, area deprivation indices have been used to predict health care outcomes. This deprivation score is an aggregate score developed from US Census data (58), which can be linked to individual-level data to provide some estimates of cluster effects. It has been successfully used to assist in the prediction of hospital readmission (35) and outcomes in hospitalized patients with community-acquired pneumonia (68). Investigators have used machine learning to aggregate Web search and location data, linked with restaurant data, to identify potentially unsafe restaurants (51). Aggregate data such as these have the capability to revolutionize the performance of model predictions in epidemiology.
Missing data, regardless of the mechanism creating the missingness, is an issue across all analytics. Many traditional regression models will drop cases with missing data and run the model. The majority of machine learning models will not run with missing data therefore, care to ensure data are complete is critical. One solution to this problem is data imputation, a technique to generate reasonable synthetic values at random when data are missing completely. These approaches have the opportunity to reduce error in missingness by accounting for nonlinear relationships in the imputator (42). Examples of machine learning missing data imputers are ripe in the literature, largely basing the models on random forest approaches (11, 55, 60). In epidemiology, variations on this theme have been used to impute missing data in some studies to better define the role of age-mixing patterns in HIV transmission dynamics (5), defining burnout and stress relationships among health care workers (13), health care utilization in patients with spinal cord injuries (49), and treatment completion prediction in patients with rape-onset post-traumatic stress disorder (34).
Classification or Regression?
Another decision to make in machine learning model building is to determine the type of outcome the investigator is interested in predicting. In machine learning, classification models are considered in the context of categorical labels, whereas regression models are used for continuous labels each model has different ramifications for model building. Nearly all supervised machine learning models can handle both classification and regression problems. In-depth review of common methods used in health research is beyond the scope of this review but can be found elsewhere (71).
All supervised machine learning algorithms have various hyperparameters that should be adjusted in order to provide a valid and accurate prediction. Some examples include the learning rate in neural networks, C and sigma in a support vector machine, or k in the k-nearest neighbor algorithm. The process of adjusting these hyperparameters is called tuning. Although most machine learning models have default values for each hyperparameter, it is worth the effort to optimize these parameters. To tune hyperparameters, a subset of the data is needed. There are many heuristics to determine how much data should be used for tuning these parameters, but there is no consensus. We recommend that ∼50% of the available data be randomly selected for hyperparameter tuning through cross-validation. This is only a generic heuristic and should be modified on the basis of the variation present in the features and outcome.
The rationale for utilizing a large portion of the data for hyperparameter tuning is that the optimal parameters cannot be known before running a model. An invalid model may result if investigators do not provide appropriate values (59). Several approaches to tuning have been described (8, 31). Grid search approaches are also easy to implement and allow for prespecification of a multitude of possible values for many or all the necessary hyperparameters required by the model. The limitation of grid searching for hyperparameters is the computationally intensive computation required. Since the investigator specifies a set of values for several hyperparameters in a grid search, models must be built for all combinations of values. Model tuning is critical and continues to be discussed as a salient concept in epidemiology. Tessmer and colleagues (61) showcase this with respect to improving R0 calculations in infectious disease epidemiology and dynamics.
Another consideration is specifically for classification models. In this context, the pretraining data set may need to be balanced with respect to the class label frequency. Here, the class label with the lowest frequency of cases is termed the minority class, whereas the higher frequency is termed the majority class. For many classification algorithms, having a relatively balanced outcome is critical (what this means is debated, though as close to 50/50 as possible is ideal). Imbalance in the outcome of a model is an issue when evaluating model performance statistics. If one class label has a much higher prevalence than another, predictive accuracy may look good while the model is predicting only the majority class. In this context, downsampling, upsampling, and synthetic minority oversampling technique (SMOTE) (14) sampling of the data are common approaches (see the sidebar titled Synthetic Minority Oversampling Technique for Handling Class Imbalance). Class balancing should be done only after splitting the data and should be independent of both the training and the testing data sets. Balancing methods have been utilized in many areas of epidemiology, including in cancer survivorship prediction (24, 67), groundwater contamination (38), and mesothelioma patients (17). Alghamdi and colleagues (1) used data from the Henry Ford Health system to predict incident diabetes with cardiorespiratory fitness data using SMOTE to balance the outcome.
Issue: Class imbalance in the outcome for clinical and epidemiological data sets prevents machine learning algorithms from learning accurately.
Solution: Synthetic minority oversampling technique (SMOTE) is a technique in which the minority class (e.g., the group with the lowest frequency) in a classification problem is oversampled by creating synthetic samples that are similar to actual samples. With this approach, the machine learning algorithm has more examples of the minority class from which to learn. The algorithm can further be combined with undersampling of the majority class to create more balanced outcomes in the data set.
Conclusion: The combination of SMOTE and undersampling performs better than undersampling alone because it focuses learning on the minority class.
Furthermore, continuous data typically should be normalized prior to training to standardize the scale of multiple continuous features and improve computational performance. It is important to normalize the data after splitting the data sets into pretraining, training, and testing. It is not advised to normalize and then split, as data leakage may occur, resulting in aberrant model performance statistics. The goal of the test set is to make it as independent of the other data as possible. If the cases ending up in the test set have features that have been scaled in consideration of some of the data in the training set, leakage will become an issue. Standardization and normalization of continuous data are necessary for many machine learning models. Seligman and colleagues (54) used data standardization approaches to understand social determinants of health in the Health and Retirement Study.
After identification of optimal hyperparameters, the next step is to split the remaining data into training and testing data sets. When defining the proportion of cases to use for a training data set, researchers face many considerations, but there is no proportion that should be deemed always acceptable. When selecting the proportion of cases for a training set, major considerations include (a) number of cases, (b) number of features, and (c) amount of variation in the features. The importance lies with how well one's training data set describes all the possible patterns of data and their potential prediction of the label. In the literature reviewed, 80% of the cases are most often used for training, although this is simply a heuristic and is not evidence based.
Similar to the hyperparameter tuning set described above, training data must be balanced with respect to the outcome for many models in the context of classification. As above, they should be balanced after splitting and be independent of all data in pretraining and testing data sets. Again as described above for hyperparameter tuning, normalization or standardization of continuous variables should be conducted after splitting.
Once again, testing the performance of the tuned and trained machine learning model in a separate data set (the test set) necessitates a proportion of the total data. The goal is to have a useful representation of real life in the testing data set. One must be careful to ensure that there is no spillover from training data sets. Here, no balancing of the outcome minority class should occur however, if features are normalized for training, they should also be normalized in the testing set. Here, the normalization factors should be applied from the training set. For example, if a column in the training set is normalized such that the individual value is subtracted from the column mean and divided by the column standard deviation in the training set (a common method of normalization), the mean and standard deviation from the training set should be applied to the values in the testing data set. The rationale for this approach arises when a model is in production. In this scenario, a single row of data (i.e., an individual's data) is supplied to the model for prediction. Here, there would be no other data from which to standardize this individual, other than the training data mean and standard deviation.
Estimating Treatment Effects
Machine learning has traditionally been focused strictly on predictive modeling, at the expense of determining treatment effects. However, causal inference and treatment effect estimation are central considerations for epidemiologists. In machine learning, treatment effects have historically been of little interest because models are created to produce predictions of the future as opposed to direct interpretation of predictor–outcome relationships (e.g., average treatment effects). This approach does not translate to a model that is effective for causal inference of model parameters. However, investigators have developed several methods for estimating treatment effects from machine learning models (20).
Heterogeneous Treatment Effects
Investigators utilizing machine learning approaches have begun to explore heterogeneous treatment effects as opposed to the overall average treatment effects (2, 10, 16, 26, 30, 44). Heterogeneous treatment effects are those that are systematically different within different groups of study subjects, often called conditional average treatment effects. One can think of identifying heterogeneous treatment effects as identifying effect modification however, exploration of heterogeneous treatment effects can be much more rigorous, comprising multiple features as opposed to just a one-way or two-way interaction term in a regression model. Investigators have developed machine learning models to detect these very specific clusters of individuals who showcase different treatment effects within their cluster of similar individuals. The most prominent example of cluster detection for calculating heterogeneous treatment effects is within causal forest models, a form of random forests that allows for the detection of subgroups of similar individuals who display different predictor–outcome effects (66). These models have been sparsely used in epidemiology a 2017 example from Baum and colleagues (4) evaluates heterogeneous treatment effects in the Look AHEAD trial, an evaluation of weight loss interventions for reducing cardiovascular complications of type 2 diabetes (see the sidebar titled Heterogeneous Treatment Effects in the Look AHEAD Trial). Other methods are available for identifying these effects in machine learning, including through the use of Bayesian additive regression trees (BART) and artificial neural networks (ANNs). In 2019, Künzel and colleagues (36) presented X-learner, a unified method to calculate heterogeneous treatment effects that allows for computation in the presence of complex distributions of treatment effects.
Issue: The Look AHEAD trial found no significant reduction in cardiovascular events in type 2 diabetic patients when undergoing weight loss interventions. Therefore, the average treatment effect of the weight loss intervention was not significantly associated with cardiovascular events in type 2 diabetics.
Solution: Using a causal forest, a type of machine learning algorithm, investigators were able to identify a subset of 75% of patients enrolled, in whom the intervention was significantly associated with reductions in cardiovascular events.
Conclusion: Randomized trials, although providing the highest level of evidence, focus largely on average treatment effects. Given the varied patients enrolled, investigators may be able to identify an ineffective treatment on average. In this setting, there may be subpopulations in whom the treatments are effective or are harmful.
Defining heterogeneous treatment effects can be particularly useful in the case of a negative study, when the results are inconclusive or suggest that the intervention is not effective. In these negative studies, there may be subpopulations or clusters of individuals who will have a different and sometimes clinically meaningful treatment effect (10). Although one can use traditional methods to identify subpopulations through stratified regression modeling or the inclusion of interaction terms, the epidemiologist will run into issues with multiple testing bias, a common pitfall in frequentist statistics. In methods driven by hypothesis testing, each additional test run on the data increases the amount of statistical error present. Therefore, if an epidemiologist wishes to evaluate 10 different variables as providing different effects among the study sample, the level of statistical error increases substantially. These novel machine learning methods do not suffer from this issue because they are algorithmic approaches to defining treatment effects and not focused on hypothesis testing. Therefore, one can evaluate as many variables as desired for defining heterogeneous treatment effects without increasing statistical error. Furthermore, because these methods are not bound by the same assumptions as are frequentist statistical approaches, there is no concern for the increase in statistical error when performing repeated hypothesis tests on the same data (i.e., multiple comparisons bias).
These approaches are very timely as we move toward personalized medicine and personalized health (46). To this end, epidemiologists can have a much better analytical handle on individual-level variations in treatment effects.
Defining Model Performance
Many methods are available to define whether a trained model performs adequately to predict the outcome with little error in the testing data set. Little novelty has been observed in the area of model performance definition most epidemiologists focus on mean squared error, root mean squared error, accuracy, precision–recall, area under the receiver operating characteristic curves (AUC ROC), and F1 statistics, depending on whether the epidemiologist is modeling continuous or categorical outcomes.
All methods provide some evidence of overall model performance, though none should be considered ideal in any particular circumstance. For example, the AUC in an ROC curve is often used to define model performance. This value can be deceiving if modeling a categorical outcome that is imbalanced. In this case, an acceptable AUC can be obtained even when the model predicts only the majority class and predicts next to zero of the minority class. In this context, the early retrieval rate from the ROC curve or precision–recall curves can be used (52). The early retrieval rate can be obtained from the left-most area (generally 1-specificity of greater than 80%) of the ROC curve. Here, if the model predicts only the majority class, one will see a low AUC because the model has a low sensitivity where the false positive rate is low. This result suggests that the model does not predict positive cases (typically the minority class). Care must be taken, when evaluating any classification model, to ensure that if the labels are not balanced then appropriate measures of model performance are used.
Explaining the Model
Machine learning algorithms are often called black boxes because most of these predictive models are difficult to explain for a single individual. Here, the practitioner is forced to believe the model, without any understanding of the potential for false positives or negatives for an individual prediction. For example, if a classification model that has an 85% balanced accuracy is being used in practice (e.g., accounting for any imbalance), it will still be wrong 15% of the time. If a new data point is supplied to the model and provides a prediction, the practitioner cannot understand whether this particular prediction is more or less likely to be in error. To correct this deficiency, novel approaches such as the local interpretable model-agnostic explanations (LIME) tests have been developed (48) and successfully used in epidemiologic studies. For example, Pereira and colleagues (43) used LIME to unbox a random forest classifier to enhance local interpretation of features in brain lesion research.
Allan & Barbara Pease
who have good eyesight but who cannot see.
Published in Australia by
Copyright © Allan Pease 2004
The right of Allan and Barbara Pease to be identified as the authors of this work has been asserted by them in accordance
with the Copyright, Designs and Patents Act 1988.
Printed in Australia by McPherson's Printing Group
Art direction by Santamaria www.santamaria.co.uk
Illustrations by Piero and John Hepworth
Cover design by Sarah Baron, Harper Collins Publishers Australia
All Things Are Not What They Seem
How Well Do You Know the Back of Your Hand?
How Well Can You Spot Body Language Contradictions? How We Wrote This Book
Your Body Language Dictionary
1. Understanding the Basics 7
How Body Language Reveals Emotions and Thoughts Why Women are More Perceptive
How Fortune-Tellers Know So Much Inborn, Genetic or Learned Culturally? Some Basic Origins
Three Rules for Accurate Reading Why It Can be Easy to Misread Why Kids are Easier to Read Can You Fake it?
True-Life Story: The Lying Job Applicant How to Become a Great Reader
2. The Power Is in Your Hands 31
The Law of Cause and Effect Palm Power Our Audience Experiment
An Analysis of Handshake Styles Who Should Reach First?
How Dominance and Control Are Communicated The Submissive Handshake
How to Create Equality How to Create Rapport
How to Disarm a Power Player The Cold, Clammy Handshake Gaining the Left Side Advantage When Men and Women Shake Hands The Double-Hander
The Blair—Bush Power Game The Solution
The World's Eight Worst Handshakes The Arafat-Rabin Handshake
3. The Magic of Smiles and Laughter 66 Smiling Is a Submission Signal
Why Smiling Is Contagious How a Smile Tricks the Brain Practising the Fake Smile Smugglers Smile Less
Five Common Types of Smiles Why Laughter Is the Best Medicine Why You Should Take Laughter Seriously Why We Laugh and Talk, But Chimps Don't How Humour Heals
Laughing Till You Cry How Jokes Work The Laughter Room
Smiles and Laughter Are a Way of Bonding Humour Sells
Smiling Advice For Women Laughter In Love
Why Crossed Arms Can be Detrimental Yes. But I'm Just 'Comfortable'
Gender Differences Crossed-Arms-on-Chest
Reinforced Arm-Crossing Arm-Gripping
The Boss vs The Staff Getting the Thumbs-Up Hugging Yourself
How the Rich and Famous Reveal their Insecurity The Coffee Cup Barrier
The Power of Touch Touch their Hand Too Summary
5. Cultural Differences 107
We Were Having Pizza at the Time Take the Cultural Test
Why We're All Becoming American
Cultural Basics are the Same Almost Everywhere Greeting Differences
When One Culture Encounters Another The English Stiff-Upper-Lip
'You Dirty, Disgusting Pig!' - Nose Blowing
The Three Most Common Cross-Cultural Gestures To Touch or Not to Touch?
How to Offend Other Cultures Summary
6. Hand and Thumb Gestures 125
How the Hands Talk On the One Hand.
On the Other Hand, Gestures Improve Recall Rubbing the Palms Together
Thumb and Finger Rub Hands Clenched Together The Steeple
Holding Hands Behind the Back Thumb Displays
7. Evaluation and Deceit Signals 142 Lying Research
The Three Wise Monkeys How the Face Reveals the Truth
Women Lie the Best and That's the Truth Why It's Hard to Lie
Eight of the Most Common Lying Gestures Evaluation and Procrastination Gestures The Lying Interviewee
Chin Stroking Stalling Clusters
Head Rubbing and Slapping Gestures Why Bob Always Lost at Chess The Double Meaning
8. Eye Signals 165 The Dilating Pupils Take the Pupil Test
Women Are Better at It, as Usual Giving Them the Eye
The 'Looking Up' Cluster How Men's Fires Get Lit
Gaze Behaviour - Where Do You Look? How to Keep Eye Contact in a Nudist Colony How to Grab a Man's Attention
Most Liars Look You in the Eye
How to Avoid being Attacked or Abused The Sideways Glance
Extended Blinking Darting Eyes
The Geography of the Face The Politician's Story
Look Deep Into My Eyes, Baby The First 20 Seconds of an Interview What Channel Are You Tuned to?
How to Hold Eye Contact with an Audience How to Present Visual Information
9. Space Invaders - Territories and Personal Space 192 Personal Space
Practical Applications of Zone Distances Who Is Moving In on Whom?
Why We Hate Riding in Lifts Why Mobs Become Angry Spacing Rituals
Cultural Factors Affecting Zone Distances Why Japanese Always Lead When They Waltz Country vs. City Spatial Zones
Territory and Ownership Car Territory
10. How the Legs Reveal What the Mind Wants to Do 209
Everybody's Talking About a New Way of Walking How Feet Tell the Truth
The Four Main Standing Positions Defensive, Cold or 'Just Comfortable'? How We Move from Closed to Open The European Leg Cross
When the Body Closes, so Does the Mind Figure Four Leg Clamp
The Short Skirt Syndrome The Leg Twine
Put Your Right Foot In, Put Your Right Foot Out Summary
11. The 13 Most Common Gestures You'll See Daily 229
Why You Should Learn to Nod How to Encourage Agreement The Head Shake
The Basic Head Positions The Head Shrug
How We Show We're Ready for Action The Cowboy Stance
Sizing Up the Competition The Legs-Spread
Leg-Over-the-Arm-of-Chair Straddling a Chair
Gestures That Show When a Person is Ready The Starter's Position
12. Mirroring - How We Build Rapport 250
Creating the Right Vibes Mirroring on a Cellular Level
Mirroring Differences Between Men and Women What to Do About It if You're Female
When Men and Women Start to Look Alike Do We Resemble Our Pets?
Monkey See, Monkey Do Matching Voices
Intentionally Creating Rapport Who Mirrors Whom?
13. The Secret Signals of Cigarettes, Glasses and Make-up 265
Differences Between Men and Women Smoking as a Sexual Display
How to Spot a Positive or Negative Decision Cigar Smokers
How Smokers End a Session How to Read Glasses
Peering-Over-the-Glasses Wearing Glasses on the Head The Power of Glasses and Make-up A Little Lippy, Lady?
Briefcase Signals Summary
14. How the Body Points to Where the Mind Wants to Go 279
What Body Angles Say How We Exclude Others Seated Body Pointing Foot Pointing
15. Courtship Displays and Attraction Gestures 286 The Emergence of the Colourful Male
Why Women Always Call the Shots Differences Between Men and Women The Attraction Process
The 13 Most Common Female Courtship Gestures and Signals
How Beautiful People Miss Out
What Men Look At in Women's Bodies Is He a Bum, Boobs or Leg Man? Male Courtship Signals and Gestures
Men's Bodies - What Turns Women On the Most Is She a Chest, Legs or Bum Gal?
16. Ownership, Territory and Height Signals 317 Body Lowering and Status
He's a Big Man Around Town
Why Some People Seem Taller on TV Try the Floor Test
Body Lowering and Status He's a Big Man Around Town
, Why Some People Seem Taller on TV Try the Floor Test
How Body Lowering Can Sometimes Raise Status How TV Politicians Can Win Votes
How to Placate Angry People What's Love Got To Do With It?
Some Strategies For Gaining Perceived Height Summary
17. Seating Arrangements - Where to Sit and Why 330 Take the Table Test
King Arthur's Concept Keeping Two People Involved Rectangular Board Tables
Why Teacher's Pet Sits On the Left Power Plays at Home
How to Make an Audience Cry The Attention Zone
An Experiment in Learning Getting a Decision Over Dinner Summary
18. Interviews, Power Plays and Office Politics 346
Why James Bond Looked Cool, Calm and Collected The Nine Golden Keys to Making Great First Impressions When Someone Keeps You Waiting
Seven Simple Strategies for Giving You the Extra Edge Summary
How to Switch Table Territories Seated Body Pointing
How to Re-arrange an Office Summary
19. Putting It All Together 364
How Well Can You Read Between the Lines? How Did You Rate?
The Six Secrets of Attractive Body Language
These are some of the people who have directly or indirectly contributed to this book, whether they knew it or not:
Dr John Tickel, Dr Dennis Waitley, Dr Andre Davril, Professor Phillip Hunsaker, Trevor Dolby, Armin Gontermann, Lothar Menne, Ray & Ruth Pease, Malcolm Edwards, Ian Marshall, Laura Meehan, Ron 6c Toby Hale, Darryl Whitby, Susan Lamb, Sadaaki Hayashsi, Deb Mehrtens, Deb Hinckesman, Doreen Carroll, Steve Wright, Derryn Hinch, Dana Reeves, Ronnie Corbett, Vanessa Feltz, Esther Rantzen, Jonathan Coleman, Trish Goddard, Kerri-Anne Kennerley, Bert Newton, Roger Moore, Lenny Henry, Ray Martin, Mike Walsh, Don Lane, Ian Lesley, Anne Diamond, Gerry & Sherry Meadows, Stan Zermarnik, Darrel Somers, Andres Kepes, Leon Byner, Bob Geldof, Vladimir Putin, Andy McNab, John Howard, Nick & Katherine Greiner, Bryce Court-ney, Tony & Cherie Blair, Greg & Kathy Owen, Lindy Chamberlain, Mike Stoller, Gerry & Kathy Bradbeer, Ty & Patti Boyd, Mark Victor Hansen, Brian Tracy, Kerry Packer, Ian Botham, Helen Richards, Tony Greig, Simon Townsend, Diana Spencer, Princes William and Harry, Prince Charles, Dr Desmond Morris, Princess Anne, David & Jan Goodwin, Iven Frangi, Victoria Singer, John Nevin, Richard Otton, Raoul Boielle, Matthew Braund, Doug Constable, George Deveraux, Rob Edmonds, Gerry Hatton, John Hepworth, Bob Heussler, Gay Huber, Ian McKillop, Delia Mills, Pamela Anderson, Wayne Mugridge, Peter Opie, David Rose, Alan White, Rob Winch, Ron Tacchi, Barry Markoff, Christine Maher, Sallie & Geoff Burch, John Fenton, Norman & Glenda Leonard
By a man's fingernails, by his coat-sleeve, by his boots, by his trouser-knees, by the calluses of his forefinger and
thumb, by his expression, by his shirt-cuffs, by his move-ments - by each of these things a man's calling is plainly
revealed. That all united should fail to enlighten the com-petent enquirer in any case is almost inconceivable.
I joined the life insurance business at the age of twenty, and went on to break several sales records for the firm I worked for, becoming the youngest person to sell over a million dollars' worth of business in my first year. This achievement qualified me for the prestigious Million Dollar Round Table in the USA. As a young man I was fortunate that the techniques I'd learned as a boy in reading body language while selling pots and pans could be transferred to this new area, and was directly related to the success I could have in any venture involving people.
All Things Are Not What They Seem
The ability to work out what is really happening with a person is simple - not easy, but simple. It's about matching what you see and hear in the environment in which it all happens and drawing probable conclusions. Most people, however, only see the things they think they are seeing.
Here's a story to demonstrate the point:
Two men were walking through the woods when they came across a big deep hole.
'Wow . that looks deep,' says one. 'Let's toss a few pebbles in and see how deep it is.'
They threw in a few pebbles and waited, but there was no sound.
'Gee - that is a really deep hole. Let's throw one of these big rocks in. That should make a noise.'
They picked up two football-sized rocks and tossed them into the hole and waited, but still they heard nothing.
'There's a railway sleeper over here in the weeds,' said one. 'If we toss that in, it's definitely going to make some noise.' They dragged the heavy sleeper over to the hole and heaved it in, but not a sound came from the hole.
it leaped into the air and disappeared into the hole. The two men stood there, astonished at what they'd just seen.
Out of the woods came a farmer who said, 'Hey! Did you guys see my goat?'
'You bet we did! It was the craziest thing we've ever seen! It came running like the wind out of the woods and jumped into that hole!'
'Nah,' says the farmer. 'That couldn't have been my goat. My goat was chained to a railway sleeper!'
How Well Do You Know the Back
Of Your Hand?
Most men and nearly half of all women don't know what they look like from the neck down.
How Well Can You Spot Body Language
People everywhere have developed a fascination with the body language of politicians because everyone knows that politi-cians sometimes pretend to believe in something that they don't believe in, or infer that they are someone other than who they really are. Politicians spend much of their time ducking, dodging, avoiding, pretending, lying, hiding their emotions and feelings, using smokescreens or mirrors and waving to imaginary friends in the crowd. But we instinctively know that they will eventually be tripped up by contradictory body lan-guage signals, so we love to watch them closely, in anticipation of catching them out.
What signal alerts you that a politician is lying? His lips are moving.
women to completely miss the change, not only in body lan-guage but in the appearance of a completely new person! Unless you have an innate ability or have learned to read body language, the chances are you're missing most of it too. This book will show you what you've been missing.
How We Wrote This Book
Barbara and I have written The Definitive Book of Body
Lan-guage using my original book, Body LanLan-guage, as our base.
Not only have we considerably expanded on that one, we have also introduced research from new scientific disciplines, such as evolutionary biology and evolutionary psychology as well as technologies such as Magnetic Resonance Imaging (MRI), which shows what happens in the brain. We have written it in a style that means it can be opened and read on almost any page at random. We've kept the content mainly to the move-ment of the body, expressions and gestures, because these are the things you need to learn to get the most out of every face-to-face encounter. The Definitive Book of Body Language will make you more aware of your own non-verbal cues and signals, and will show you how to use them to communicate effectively and how to get the reactions you want.
This book isolates and examines each component of body language and gesture in simple terms to make it accessible to you. Few gestures are made in isolation of others however, so we have also, at the same time, tried to avoid oversimplifying things.
works makes living with it easier, whereas ignorance and lack of understanding promote fear and superstition and make us more critical of others. A birdwatcher doesn't study birds so that he can shoot them down and keep them as trophies. In the same way, the knowledge and skills in body language serve to make every encounter with another person an exciting experi-ence.
For the purpose of simplicity, and unless otherwise stated, the use of 'he' or 'him' will apply equally to both genders.
Your Body Language Dictionary
The original book was intended as a working manual for sales people, managers, negotiators and executives, but this one can be used for any aspect of your life, be it at home, on a date or at work. The Definitive Book of Body Language is the result of over 30 years of our cumulative knowledge and involvement in this field and we give you the basic 'vocabulary' you need to read attitudes and emotions. This book will give you answers to some of the most puzzling questions you've ever had about why people use some of the behaviours they do, and it will change forever your own behaviour. It will seem as if you've always been in a dark room and, while you could always feel the furnishings, the wall hangings and the door, you've never actually seen what they look like. This book will be like turning on the lights to see what was always there. But now, you'll know exactly what things are, where they are and what to do about them.
This is 'good' to Westerners, 'one' to Italians, 'five' to Japanese and 'up yours' to the Greeks
Everyone knows someone who can walk into a room full of people and, within minutes, give an accurate description about the relationships between those people and what they are feeling. The ability to read a person's attitudes and thoughts by their behaviour was the original communication system used by humans before spoken language evolved.
Today's politicians understand that politics is about image and appearance and most high-profile politicians now have personal body language consultants to help them come across as being sincere, caring and honest, especially when they're not.
It seems almost incredible that, over the thousands of years of our evolution, body language has been actively studied on any scale only since the 1960s and that most of the public has become aware of its existence only since our book Body
Lan-guage was published in 1978. Yet most people believe that
speech is still our main form of communication. Speech has been part of our communication repertoire only in recent times in evolutionary terms, and is mainly used to convey facts and data. Speech probably first developed between 2 million and 500,000 years ago, during which time our brain tripled its size. Before then, body language and sounds made in the throat were the main forms of conveying emotions and feel-ings, and that is still the case today. But because we focus on the words people speak, most of us are largely uninformed about body language, let alone its importance in our lives.
Our spoken language, however, recognises how important body language is to our communication. Here are just a few of the phrases we use —
Get it off your chest. Keep a stiff upper lip. Stay at arm's length. Keep your chin up.
Shoulder a burden. Face up to it. Put your best foot forward. Kiss my butt
In the Beginning.
Silent movie actors like Charlie Chaplin were the pioneers of body language skills, as this was the only means of communi-cation available on the screen. Each actor's skill was classed as good or bad by the extent to which he could use gestures and body signals to communicate to the audience. When talking films became popular and less emphasis was placed on the non-verbal aspects of acting, many silent movie actors faded into obscurity and only those with good verbal and non-verbal skills survived.
As far as the academic study of body language goes, perhaps the most influential pre-twentieth-century work was Charles Darwin's The Expression of the Emotions in Man and
Animals, published in 1872, but this work tended to be read
mainly by academics. However, it spawned the modern studies of facial expressions and body language, and many of Darwin's ideas and observations have since been validated by researchers around the world. Since that time, researchers have noted and recorded almost a million non-verbal cues and signals. Albert Mehrabian, a pioneer researcher of body lan-guage in the 1950s, found that the total impact of a message is about 7% verbal (words only) and 38% vocal (including tone of voice, inflection and other sounds) and 55% non-verbal.
It's how you looked when you said it, not what you actually said.
Like Mehrabian, he found that the verbal component of a face-to-face conversation is less than 35% and that over 65% of communication is done non-verbally. Our analysis of thou-sands of recorded sales interviews and negotiations during the 1970s and 1980s showed that, in business encounters, body language accounts for between 60 and 80% of the impact made around a negotiating table and that people form 60 to 80% of their initial opinion about a new person in less than four minutes. Studies also show that when negotiating over the telephone, the person with the stronger argument usually wins, but this is not so true when negotiating face-to-face, because overall we make our final decisions more on what we see than what we hear.
Why It's Not What You Say
Despite what it may be politically correct to believe, when we meet people for the first time we quickly make judgements about their friendliness, dominance and potential as a sexual partner - and their eyes are not the first place we look.
Most researchers now agree that words are used primarily for conveying information, while body language is used for negotiating interpersonal attitudes and in some cases is used as a substitute for verbal messages. For example, a woman can give a man a 'look to kill' and will convey a very clear message to him without opening her mouth.
Regardless of culture, words and movements occur together with such predictability that Birdwhistell was the first to claim that a well-trained person should be able to tell what move-ment a person is making by listening to their voice. Birdwhistell even learned how to tell what language a person was speaking, simply by watching their gestures.
Many people find difficulty in accepting that humans are still biologically animals. We are a species of primate — Homo
sapiens — a hairless ape that has learned to walk on two limbs
are still dominated by biological rules that control our actions, reactions, body language and gestures. The fascinating thing is that the human animal is rarely aware that its postures, move-ments and gestures can tell one story while its voice may be telling another.
How Body Language Reveals Emotions and
Body language is an outward reflection of a person's emo-tional condition. Each gesture or movement can be a valuable key to an emotion a person may be feeling at the time. For example, a man who is self-conscious about gaining weight may tug at the fold of skin under his chin the woman who is aware of extra pounds on her thighs may smooth her dress down the person who is feeling fearful or defensive might fold their arms or cross their legs or both and a man talking with a large-breasted woman may consciously avoid staring at her breasts while, at the same time, unconsciously use groping ges-tures with his hands.
Prince Charles finds a bosom buddy
The key to reading body language is being able to understand a person's emotional condition while listening to what they are saying and noting the circumstances under which they are saying it. This allows you to separate fact from fiction and reality from fantasy. In recent times, we humans have had an obsession with the spoken word and our ability to be conver-sationalists. Most people, however, are remarkably unaware of body language signals and their impact, despite the fact that we now know that most of the messages in any face-to-face conversation are revealed through body signals. For example, France's President Chirac, USA's President Ronald Reagan and Australia's Prime Minister Bob Hawke all used their hands to reveal the relative sizes of issues in their mind. Bob Hawke once defended pay increases for politicians by comparing their salaries to corporate executive salaries. He claimed that exec-utive salaries had risen by a huge amount and that proposed politicians' increases were relatively smaller. Each time he mentioned politicians' incomes, he held his hands a yard (1m) apart. When he mentioned executive salaries, however, he held them only a foot (30cm) apart. His hand distances revealed that he felt politicians were getting a much better deal than he was prepared to admit.
President Jacques Chirac — measuring the size of an issue or simply boasting about his love life?
Why Women are More Perceptive
we say someone is 'perceptive' or 'intuitive' about people, we are unknowingly referring to their ability to read another person's body language and to compare these cues with verbal signals. In other words, when we say that we have a 'hunch' or 'gut feeling' that someone has told us a lie, we usually mean that their body language and their spoken words don't agree. This is also what speakers call audience aware-ness, or relating to a group. For example, if an audience were sitting back in their seats with their chins down and arms crossed on their chest, a 'perceptive' speaker would get a hunch or feeling that his delivery was not going across well. He would realise that he needed to take a different approach to gain audience involvement. Likewise, a speaker who was not 'perceptive' would blunder on regardless.
Being 'perceptive' means being able to spot the contradictions between someone's words and their body language.
Overall, women are far more perceptive than men, and this has given rise to what is commonly referred to as 'women's intu-ition'. Women have an innate ability to pick up and decipher non-verbal signals, as well as having an accurate eye for small details. This is why few husbands can lie to their wives and get away with it and why, conversely, most women can pull the wool over a man's eyes without his realising it.
men also scored well. Female intuition is particularly evident in women who have raised children. For the first few years, the mother relies almost solely on the non-verbal channel to com-municate with the child and this is why women are often more perceptive negotiators than men because they practise reading signals early.
What Brain Scans Show
Most women have the brain organisation to out-communicate any man on the planet. Magnetic Resonance Imaging brain scans (MRI) clearly show why women have far greater capac-ity for communicating with and evaluating people than men do. Women have between fourteen and sixteen areas of the brain to evaluate others' behaviour versus a man's four to six areas. This explains how a woman can attend a dinner party and rapidly work out the state of the relationships of other couples at the party - who's had an argument, who likes who and so on. It also explains why, from a woman's standpoint, men don't seem to talk much and, from a man's standpoint, women never seem to shut up.
As we showed in Why Men Don't Listen & Women Can't
Read Maps (Orion), the female brain is organised for
multi-tracking — the average woman can juggle between two and four unrelated topics at the same time. She can watch a television programme while talking on the telephone plus listen to a second conversation behind her, while drinking a cup of coffee. She can talk about several unrelated topics in the one conversation and uses five vocal tones to change the subject or emphasise points. Unfortunately, most men can only identify three of these tones. As a result, men often lose the plot when women are trying to communicate with them.
Studies show that a person who relies on hard visual evi-dence face to face about the behaviour of another person is more likely to make more accurate judgements about that person than someone who relies solely on their gut feeling.
The evidence is in the person's body language and, while women can do it subconsciously, anyone can teach themselves consciously to read the signals. That's what this book is about.
How Fortune-Tellers Know So Much
The fortune-teller gazed into her crystal ball and then started laughing uncontrollably. So John punched her on the nose. It was the first time he'd ever struck a happy medium.
To demonstrate the point, here now is a psychic reading for you personally. Imagine you've come to a dimly lit, smoke-filled room where a jewel-encrusted psychic wearing a turban is seated at a low, moon-shaped table with a crystal ball:
I'm glad you've come to this session and I can see you have things that are troubling you because I am receiving strong signals from you. I sense that the things you really want out of life sometimes seem unrealistic and you often wonder whether you can achieve them. I also sense that at times you are friendly, social and outgoing to others, but that at other times you are withdrawn, reserved and cautious. You take pride in being an independent thinker but also know not to accept what you see and hear from others, without proof. You like change and variety but become restless if controlled by restrictions and routine. You want to share your innermost feelings with those closest to you but have found it unwise to be too open and revealing. A man in your life with the initial 'S' is exerting a strong influence over you right now and a woman who is born in November will contact you in the next month with an exciting offer. While you appear disciplined and controlled on the outside, you tend to be concerned and worried on the inside and at times you wonder whether or not you have made the right choice or decision.
the dog! We won't encourage you to become a fortune-teller but you'll soon be able to read others as accurately as they do.
Inborn, Genetic or Learned Culturally?
When you cross your arms on your chest, do you cross left over right or right over left? Most people cannot confidently describe which way they do this until they try it. Cross your arms on your chest right now and then try to quickly reverse the position. Where one way feels comfortable, the other feels completely wrong. Evidence suggests that this may well be a genetic gesture that cannot be changed.
Seven out of ten people cross their left arm over their right.
Much debate and research has been done to discover whether non-verbal signals are inborn, learned, genetically transferred or acquired in some other way. Evidence has been collected from observation of blind people (who could not have learned non-verbal signals through a visual channel), from observing the gestural behaviour of many different cultures around the world and from studying the behaviour of our nearest anthro-pological relatives, the apes and monkeys.
emotion, which led them to the conclusion that these gestures must also be inborn.
Cultural differences are many but the basic body language signals are the same everywhere.
Debate still exists as to whether some gestures are culturally learned, and become habitual, or are genetic. For example, most men put on a coat right arm first most women put it on left arm first. This shows that men use their left brain hemi-sphere for this action while women use the right hemihemi-sphere. When a man passes a woman in a crowded street, he usually turns his body towards her as he passes she instinctively turns her body away from him to protect her breasts. Is this an inborn female reaction or has she learned to do this by uncon-sciously watching other females?
Some Basic Origins
Most of the basic communication signals are the same all over the world. When people are happy they smile when they are sad or angry they frown or scowl. Nodding the head is almost universally used to indicate 'yes' or affirmation. It appears to be a form of head lowering and is probably an inborn gesture because it's also used by people born blind. Shaking the head from side to side to indicate 'no' or negation is also universal and appears to be a gesture learned in infancy. When a baby has had enough milk, it turns its head from side to side to reject its mother's breast. When the young child has had enough to eat, he shakes his head from side to side to stop any attempt to spoon-feed him and, in this way, he quickly learns to use the head shaking gesture to show disagreement or a neg-ative attitude.
The head-shaking gesture signals 'no' and owes its origin to breastfeeding.
The evolutionary origin of some gestures can be traced to our primitive animal past. Smiling, for example, is a threat gesture for most carnivorous animals, but for primates it is done in conjunction with non-threatening gestures to show submis-sion.
Baring the teeth and nostril flaring are derived from the act of attacking and are primitive signals used by other primates. Sneering is used by animals to warn others that, if necessary, they'll use their teeth to attack or defend. For humans, this gesture still appears even though humans won't usually attack with their teeth.
Human and animal sneering — you wouldn't want to go on a date with either of these two
Nostril flaring allows more air to oxygenate the body in prepa-ration for fight or flight and, in the primate world, it tells others that back-up support is needed to deal with an immi-nent threat. In the human world, sneering is caused by anger, irritation, when a person feels under physical or emotional threat or feels that something is not right.
The Shoulder Shrug is also a good example of a universal gesture that is used to show that a person doesn't know or doesn't understand what you are saying. It's a multiple gesture that has three main parts: exposed palms to show nothing is being concealed in the hands, hunched shoulders to protect the throat from attack and raised brow which is a universal, sub-missive greeting.
The Shoulder Shrug shows submission
Just as verbal language differs from culture to culture, so some body language signals can also differ. Whereas one gesture may be common in a particular culture and have a clear inter-pretation, it may be meaningless in another culture or even have a completely different meaning. Cultural differences will be covered later, in Chapter 5.
Three Rules for Accurate Reading
What you see and hear in any situation does not necessarily reflect the real attitudes people may actually have. You need to follow three basic rules to get things right.
Rule 1. Read Gestures in Clusters
One of the most serious errors a novice in body language can make is to interpret a solitary gesture in isolation of other ges-tures or circumstances. For example, scratching the head can mean a number of things - sweating, uncertainty, dandruff, fleas, forgetfulness or lying - depending on the other gestures that occur at the same time. Like any spoken language, body language has words, sentences and punctuation. Each gesture is like a single word and one word may have several different meanings. For example, in English, the word 'dressing' has at least ten meanings including the act of putting on clothing, a sauce for food, stuffing for a fowl, an application for a wound, fertiliser and grooming for a horse.
It's only when you put a word into a sentence with other words that you can fully understand its meaning. Gestures come in 'sentences' called clusters and invariably reveal the truth about a person's feelings or attitudes. A body language cluster, just like a verbal sentence, needs at least three words in it before you can accurately define each of the words. The 'per-ceptive' person is the one who can read the body language sentences and accurately match them against the person's verbal sentences.
Scratching the head can mean uncertainty but it's also a sign of dandruff.
So always look at gesture clusters for a correct reading. Each of us has one or more repetitive gestures that simply reveal we are either bored or feeling under pressure. Continual hair touching or twirling is a common example of this but, in iso-lation of other gestures, it's likely to mean the person is feeling uncertain or anxious. People stroke their hair or head because that's how their mother comforted them when they were chil-dren.
To demonstrate the point about clusters, here's a common
Critical Evaluation gesture cluster someone might use when
they are unimpressed with what they are hearing:
You're losing points with this man
The main Critical Evaluation signal is the hand-to-face gesture, with the index finger pointing up the cheek while another finger covers the mouth and the thumb supports the chin. Further evidence that this listener is having critical thoughts about what he hears is supported by the legs being tightly crossed and the arm crossing the body (defensive) while the head and chin are down (negative/hostile). This body lan-guage 'sentence' says something like, 'I don't like what you're saying', 'I disagree' or 'I'm holding back negative feelings'.
Hillary Clinton uses this cluster when she's not convinced
Rule 2. Look for Congruence
Research shows that non-verbal signals carry about five times as much impact as the verbal channel and that, when the two are incongruent people - especially women - rely on the non-verbal message and disregard the non-verbal content.
If you, as the speaker, were to ask the listener shown above to give his opinion about something you've said and he replied that he disagreed with you, his body language signals would be congruent with his verbal sentences, that is, they would match. If, however, he said he agreed with what you said, he would more likely be lying because his words and gestures would be incongruent.
When a person's words and body language
are in conflict, women ignore what is said.
If you saw a politician standing behind a lectern speaking con-fidently but with his arms tightly folded across his chest (defensive) and chin down (critical/hostile), while telling his audience how receptive and open he is to the ideas of young people, would you be convinced? What if he attempted to con-vince you of his warm, caring approach while giving short, sharp karate chops to the lectern? Sigmund Freud once reported that while a patient was verbally expressing happi-ness with her marriage, she was unconsciously slipping her wedding ring on and off her finger. Freud was aware of the sig-nificance of this unconscious gesture and was not surprised when marriage problems began to surface.
Observation of gesture clusters and congruence of the verbal and body language channels are the keys to accurately interpreting attitudes through body language.
Rule 3. Read Gestures in Context
All gestures should be considered in the context in which they occur. If, for example, someone was sitting at a bus terminal
with his arms and legs tightly crossed and chin down and it was a cold winter's day, it would most likely mean that he was cold, not defensive. If, however, the person used the same ges-tures while you were sitting across a table from him trying to sell him an idea, product or service, it could be correctly inter-preted as meaning that the person was feeling negative or rejecting your offer.
Throughout this book all body language gestures will be con-sidered in context and, where possible, gesture clusters will be examined.
Why It Can be Easy to Misread
Someone who has a soft or limp handshake — especially a man — is likely to be accused of having a weak character and the next chapter on handshake techniques will explore the reason behind this. But if someone has arthritis in their hands it is likely that
they will also use a soft handshake to avoid the pain of a strong one. Similarly, artists, musicians, surgeons and those whose occupation is delicate and involves use of their hands generally prefer not to shake hands, but, if they are forced into it, they may use a 'dead fish' handshake to protect their hands.
Someone who wears ill-fitting or tight clothing may be unable to use certain gestures, and this can affect their use of body lan-guage. For example, obese people can't cross their legs. Women who wear short skirts will sit with their legs tightly crossed for protection, but this results in them looking less approachable and less likely to be asked to dance at a nightclub. These cir-cumstances apply to the minority of people, but it is important to consider what effect a person's physical restrictions or dis-abilities may have on their body movement.
Why Kids are Easier to Read
Older people are harder to read than younger ones because they have less muscle tone in the face.
The speed of some gestures and how obvious they look to others is also related to the age of the individual. For example, if a five-year-old child tells a lie, he's likely to immediately cover his mouth with one or both hands.
The act of covering the mouth can alert a parent to the lie and this mouth-covering gesture will likely continue throughout
the person's lifetime, usually only varying in the speed at which it's done. When a teenager tells a lie, the hand is brought to the mouth in a similar way to the five-year-old, but instead of the obvious hand-slapping gesture over the mouth, the fingers rub lightly around it.
The teenager telling a lie
The original mouth-covering gesture becomes even faster in adulthood. When an adult tells a lie, it's as if his brain instructs his hand to cover his mouth in an attempt to block the deceitful words, just as it did for the five-year-old and the teenager. But, at the last moment, the hand is pulled away from the face and a nose touch gesture results. This is simply an adult's version of the mouth-covering gesture that was used in childhood.
Bill Clinton answering questions about Monica Lewinsky in front of the Grand Jury
This shows how, as people get older, their gestures become more subtle and less obvious and is why it's often more diffi-cult to read the gestures of a fifty-year-old than those of a five-year-old.
Can You Fake it?
We are regularly asked, 'Can you fake body language?' The general answer to this question is 'no', because of the lack of congruence that is likely to occur between the main gestures, the body's micro-signals and the spoken words. For example, open palms are associated with honesty but when the faker holds his palms out and smiles at you as he tells a lie, his micro-gestures give him away. His pupils may contract, one eyebrow may lift or the corner of his mouth may twitch, and these signals contradict the open palm gesture and the sincere smile. The result is that the receivers, especially women, tend not to believe what they hear.
Body language is easier to fake with men than with women because, overall, men
aren't good readers of body language.
True-Life Story:The Lying Job Applicant
we were interviewing a man who was explaining why he had quit his last job. He told us that there had been insufficient future opportunity available to him and that it was a hard decision to leave as he got on well with all the staff there. A female interviewer said she had an 'intuitive feeling' that the applicant was lying and that he had negative feelings about his forrner boss, despite the applicant's continual praising of his boss. During a review of the interview on slow-motion video, we noticed that each time the applicant mentioned his former
boss a split-second sneer appeared on the left side of his face. Often these contradictory signals will flash across a person's face in a fraction of a second and are missed by an untrained observer. We telephoned his former boss and discovered the applicant had been fired for dealing drugs to other staff members. As confidently as this applicant had tried to fake his body language, his contradictory micro-gestures gave the game away to our female interviewer.
The key here is being able to separate the real gestures from fake ones so a genuine person can be distinguished from a liar or impostor. Signals like pupil dilation, sweating and blushing cannot be consciously faked but exposing the palms to try to appear honest is easily learned.
Fakers can only pretend for a short period of time.
There are, however, some cases in which body language is deliberately faked to gain certain advantages. Take, for example, the Miss World or Miss Universe contest, in which each contestant uses studiously learned body movements to give the impression of warmth and sincerity. To the extent that each contestant can convey these signals, she will score points from the judges. But even the expert contestants can only fake body language for a short period of time and even-tually the body will show contradictory signals that are independent of conscious actions. Many politicians are experts in faking body language in order to get the voters to believe what they are saying, and politicians who can success-fully do this - such as John F Kennedy and Adolf Hitler — are said to have 'charisma'.
In summary, it is difficult to fake body language for a long period of time but, as we will discuss, it's important to learn how to use positive body language to communicate with others and to eliminate negative body language that may give out the wrong message. This can make it more comfortable to
be with others and make you more acceptable to them, which is one of the aims of this book.
How to Become a Great Reader
Set aside at least fifteen minutes a day to study the body lan-guage of other people, as well as acquiring a conscious awareness of your own gestures. A good reading ground is anywhere that people meet and interact. An airport is a par-ticularly good place for observing the entire spectrum of human gestures as people openly express eagerness, anger, sorrow, happiness, impatience and many other emotions through body language. Social functions, business meetings and parties are also excellent. When you become proficient at the art of reading body language, you can go to a party, sit in a corner all evening and have an exciting time just watching other people's body language rituals.
Modern humans are worse at reading body signals than their ancestors because we
are now distracted by words.
Television also offers an excellent way of learning. Turn down the sound and try to understand what is happening by first watching the picture. By turning the sound up every few minutes, you will be able to check how accurate your non-verbal readings are and, before long, it will be possible to watch an entire programme without any sound and under-stand what is happening, just as deaf people do.
Learning to read body language signals not only makes you more acutely aware of how others try to dominate and manip-ulate, it brings the realisation that others are also doing the same to us and, most importantly, it teaches us to be more sen-sitive to other people's feelings and emotions.
we have now witnessed the emergence of a new kind of
social scientist - the Body Language Watcher. Just as the bird-watcher loves watching birds and their behaviour, so the Body Language Watcher delights in watching the non-verbal cues and signals of human beings. He watches them at social func-tions, at beaches, on television, at the office or anywhere that people interact. He's a student of behaviour who wants to learn about the actions of his fellow humans so that he may ultimately learn more about himself and how he can improve his relationships with others.
What's the difference between an observer and a stalker? A clipboard and pen.
THE POWER IS IN
How the Palms and Handshakes
Are used to control
In ancient times, open palms were used to show that no weapons were being concealed
It was Adam's first day on the job with his new PR company and he wanted to make a good impression on everyone. As he was introduced to colleague after colleague, he shook their hands enthusiastically and gave everyone a broad smile. Adam stood 6 foot 3 inches (1.9m) tall and was good looking, well dressed and certainly looked like a successful PR man. He always gave a firm handshake, just the way his father had taught him when he was young. So firm in fact, that it drew blood on the ring fingers of two female colleagues and left several others feeling injured. Other men competed with Adam's handshake — that's what men do. The women,
however, suffered in silence and soon were whispering, 'Stay away from that new guy Adam — he's a bruiser!' The men never brought it up - but the women simply avoided Adam. And half the firm's bosses were women.
Here's a handy thought — whether you are heavy-handed or high-handed, engage in sleight-of-hand to avoid a hand-to-mouth existence, you might have to show your hand sometimes to gain the upper hand. Don't get caught red-handed, or try to wash your hands of a mistake, because if you bite the hand that feeds you, things could get out of hand.
The hands have been the most important tools in human evolution and there are more connections between the brain and the hands than between any other body parts. Few people ever consider how their hands behave or the way they shake hands when they meet someone. Yet those first five to seven pumps establish whether dominance, submission or power plays will take place. Throughout history, the open palm has been associated with truth, honesty, allegiance and submis-sion. Many oaths are still taken with the palm of the hand over the heart, and the palm is held in the air when somebody is giving evidence in a court of law the Bible is held in the left hand and the right palm held up for the members of the court to view. One of the most valuable clues to discovering whether someone is being open and honest - or not - is to watch for palm displays. Just as a dog will expose its throat to show submission or surrender to the victor, humans use their palms to display in a similar way to show that they are unarmed and therefore not a threat.
Submissive dogs reveal their throats. Humans show their palms.
How to Detect Openness
When people want to be open or honest, they will often hold one or both palms out to the other person and say something like, 'I didn't do it!', 'I'm sorry if I upset you' or 'I'm telling you the truth'. When someone begins to open up or be truth-ful, they will likely expose all or part of their palms to the other person. Like most body language signals, this is a com-pletely unconscious gesture, one that gives you an 'intuitive' feeling or hunch that the other person is telling the truth.
The palms are intentionally used everywhere to infer an open, honest approach
When children are lying or concealing something, they'll often hide their palms behind the back. Similarly, a man who wants to conceal his whereabouts after a night out with the boys might hide his palms in his pockets, or in an arms-crossed position, when he tries to explain to his partner where he was. However, the hidden palms may give her an intuitive feeling that he is not telling the truth. A woman who is trying to hide something will try to avoid the subject or talk about a range of unrelated topics while doing various other activities at the same time.
When men lie their body language can be obvious. Women prefer to look busy as they lie.
Salespeople are taught to watch for a customer's exposed palms when he gives reasons or objections about why he can't buy a product, because when someone is giving valid reasons, they usually show their palms. When people are being open in explaining their reasons they use their hands and flash their palms whereas someone who isn't telling the truth is likely to give the same verbal responses but conceal their hands.
Keeping their hands in their pockets is a favourite ploy of men who don't want to participate in a conversation. The palms were originally like the vocal cords of body language because they did more 'talking' than any other body part and putting them away was like keeping one's mouth shut.
Palms-in-Pockets: Prince William showing the media that he doesn't want to talk
Intentional Use of the Palms to Deceive
Some people ask, 'If I tell a lie and keep my palms visible, will people be more likely to believe me?' The answer is yes - and no. If you tell an outright lie with your palms exposed, you might still appear insincere to your listeners because many of the other gestures that should also be visible when displaying honesty are absent and the negative gestures used during lying will appear and will be incongruent with the open palms. Con artists and professional liars are people who have developed the special art of making their non-verbal signals complement their verbal lies. The more effectively the professional con artist can use the body language of honesty when telling a lie, the better he is at his job.
Will you still love me when I'm old and grey?' she asked, palms visible.'Not only will I love you,' he replied, 'I'll write to you.'
The Law of Cause and Effect
It's possible, however, to appear more open and credible by practising open palm gestures when communicating with others. Interestingly, as the open palm gestures become habit-ual, the tendency to tell untruths diminishes. Most people find it difficult to lie with their palms exposed because of the law of cause and effect. If a person is being open they'll expose their palms, but just having their palms exposed makes it dif-ficult for the person to tell a convincing lie. This is because gestures and emotions are directly linked to each other. If you reel defensive, for example, you're likely to cross your arms across your chest. But if you simply cross your arms you'll begin to experience defensive feelings. And if you are talking with your palms exposed it puts even more pressure on the other person to be truthful too. In other words, open palms can help to suppress some of the false information others may tell and encourage them to be more open with you.
One of the least noticed, but most powerful, body signals is given by the human palm when giving someone directions or commands and in handshaking. When used in a certain way,
Palm Power invests its user with the power of silent authority.
There are three main palm command gestures: the Palm-Up position, the Palm-Down position and the Palm-Closed- i
Finger-Pointed position. The differences of the three positions
are shown in this example: let's say that you ask someone to pick up something and carry it to another location. We'll assume that you use the same tone of voice, the same words and facial expressions in each example, and that you change only the position of your palm.
The palm facing up is used as a submissive, non-threatening gesture, reminiscent of the pleading gesture of a street beggar and, from an evolutionary perspective, shows the person holds no weapons. The person being asked to move the item will not feel they are being pressured into it and are unlikely to feel threatened by your request. If you want someone to talk you can use the Palm-Up as a 'handover' gesture to let them know you expect them to talk and that you're ready to listen.
The Palm-Up gesture became modified over the centuries and gestures like the Single-Raised-in-the-Air, the
Palm-Over-tbe-Heart and many other variations developed.
Palm up = non threatening Palm down — authority
When the palm is turned to face downwards, you will project immediate authority. The other person will sense that you've
given them an order to move the item and may begin to feel antagonistic towards you, depending on your relationship with him or the position you have with him in a work environment.
Turning your palm from facing upwards to facing downwards completely
alters how others perceive you.
For example, if the other person was someone of equal status, he might resist a Palm-Down request and would be more likely to comply if you'd used the Palm-Up position. If the person is your subordinate, the Palm-Down gesture is seen as acceptable because you have the authority to use it.
The Nazi salute had the palm facing directly down and was the symbol of power and tyranny during the Third Reich. If Adolf Hitler had used his salute in the Palm-Up position no one would have taken him seriously - they would have laughed.
Adolf Hitler using one of history's most notable Palm-Down signals
When couples walk hand-in-hand the dominant partner, often 1 the man, walks slightly in front with his hand in the above position, palm facing backwards while she has her palm facing forward. This simple little position immediately reveals to an observer who wears the loincloths in that family.
The Palm-Closed-Finger-Pointed is a fist where the pointed finger is used like a symbolic club with which the speaker fig-uratively beats his listeners into submission. Subconsciously, it evokes negative feelings in others because it precedes a right over-arm blow, a primal move most primates use in a physical attack.
Pointing finger — 'Do it or else!'
The Palm-Closed-Finger-Pointed gesture is one of the most annoying gestures anyone can use while speaking, particularly when it beats time to the speaker's words. In some countries such as Malaysia and the Philippines, finger pointing at a person is an insult as this gesture is only used to point at animals. Malaysians will use their thumb to point to people or to give directions.
Our Audience Experiment
We conducted an experiment with eight lecturers who were asked to use each of these three hand gestures during a series of ten-minute talks to a range of audiences and we later recorded the attitudes of the participants to each lecturer. We found that the lecturers who mostly used the Palm-Up position received 84% positive testimonials from their participants,
which dropped to 52% when they delivered exactly the same presentation to another audience using mainly the Palm-Down position. The Finger-Pointed position recorded only 28% pos-itive response and some participants had walked out during the lecture.
The pointing finger creates negative feelings in most listeners
Finger pointing not only registered the least amount of posi-tive responses from the listeners they could also recall less of what the speaker had said. If you are a habitual finger-pointer, try practising the palm-up and palm-down positions and you'll find that you can create a more relaxed atmosphere and have a more positive effect on others. Alternatively, if you squeeze your fingers against your thumb to make an 'OK' type of gesture and talk using this position, you'll come across as authoritative, but not aggressive. We taught this gesture to groups of speakers, politicians and business leaders and we measured the audience reactions. The audiences who listened to the speakers who used the fingertip-touch gestures described those speakers as 'thoughtful', 'goal-oriented' and 'focused'.
Squeezing the thumb against the fingertips avoids intimidating the audience
Speakers who used the finger-pointed position were described as 'aggressive', 'belligerent' and 'rude' and recorded the lowest amount of information retention by their audience. When the speaker pointed directly at the audience, the delegates became preoccupied with making personal judgements about the speaker rather than listening to his content.
An Analysis of Handshake Styles
Shaking hands is a relic of our ancient past. Whenever primi-tive tribes met under friendly conditions, they would hold their arms out with their palms exposed to show that no weapons were being held or concealed. In Roman times, the practice of carrying a concealed dagger in the sleeve was common so for protection the Romans developed the
Lower-Arm-Grasp as a common greeting.
The Lower Arm Grasp -checking for concealed weapons — the original Roman method of greeting
The modern form of this ancient greeting ritual is the inter-locking and shaking of the palms and was originally used in the nineteenth century to seal commercial transactions between men of equal status. It has become widespread only in the last hundred years or so and has always remained in the male domain until recent times. In most Western and Euro-pean countries today it is performed both on initial greeting and on departure in all business contexts, and increasingly at parties and social events by both women and men.
The handshake evolved as a way men could
cement a commercial deal with each other.
Even in places such as Japan, where bowing is the traditional greeting, and Thailand, where they greet using the Wai — a gesture that looks similar to praying — the modern handshake is now widely seen. In most places, the hands are normally pumped five to seven times but in some countries, for example Germany, they pump two or three times with an additional hold time equal to an extra two pumps. The French are the biggest glad-handers, shaking on both greeting and departure and spending a considerable time each day shaking hands.
Who Should Reach First?
welcome them and feels forced to shake hands. Under these circumstances, salespeople are advised that it is better to wait for the other person to initiate the handshake and, if it is not forthcoming, use a small head-nod as the greeting. In some countries, shaking hands with a woman is still an uncertain practice (for example, in many Muslim countries it would be considered rude to do so instead a small head-nod is accept-able), but it's now been found that women who initiate a firm handshake are rated — in most places — as more open-minded and make better first impressions.
How Dominance and Control Are
Considering what has already been said about the impact of the Palm-Up and Palm-Down gestures, let's explore their rele-vance in handshaking.
In Roman times, two leaders would meet and greet each other with what amounted to a standing version of modern arm wrestling. If one leader was stronger than the other, his hand would finish above the other's hand in what became known as the Upper Hand position.
Let's assume that you have just met someone for the first time and you greet each other with a handshake. One of three basic attitudes is subconsciously transmitted:
1. Dominance: 'He is trying to dominate me. I'd better be cautious.'
2. Submission: 'I can dominate this person. He'll do what I want.'
3. Equality: 'I feel comfortable with this person.'
These attitudes are sent and received without our being aware of them, but they can have an immediate impact on the outcome of any meeting. In the 1970s we documented the effect of these handshake techniques in our business skills
classes and taught them as business strategies, which, with a little practice and application, can dramatically influence any face-to-face meeting, as you will see.
Dominance is transmitted by turning your hand (striped sleeve) so that your palm faces down in the handshake (see below). Your palm doesn't have to face directly down, but is the upper hand and communicates that you want to take control of the encounter.
Our study of 350 successful senior management executives (89% of whom were men) revealed that not only did almost all of the managers initiate the handshake, 88% of males and 31% of females also used the dominant handshake position. Power and control issues are generally less important to women, which probably accounts for why only one in three women attempted the Upper Hand ritual. We also found that some women will give men a soft handshake in some social contexts to imply submissiveness. This is a way of highlighting their femininity or implying that domination of her may be possible. In a business context, however, this approach can be disastrous for a woman because men will give attention to her feminine qualities and not take her seriously. Women who display high femininity in business meetings are not taken seri-ously by other business women or men, despite the fact that it's now fashionable or politically correct to say everyone is the
same. This doesn't mean a woman in business needs to act in a masculine way she simply needs to avoid signals of female-ness such as soft handshakes, short skirts and high heels if she wants equal credibility.
Women who show high feminine signals in a serious business meeting lose credibility.
In 2001, William Chaplin at the University of Alabama con-ducted a study into handshakes and found that extroverted types use firm handshakes while shy, neurotic personalities don't. Chaplin also found that women who are open to new ideas used firm handshakes. Men used the same handshakes whether they were open to new ideas or not. So it makes good business sense for women to practise firmer handshaking, par-ticularly with men.
The Submissive Handshake
The opposite of the dominant handshake is to offer your hand (striped sleeve) with your palm facing upwards (as below), symbolically giving the other person the upper hand, like a dog exposing its throat to a superior dog.
This can be effective if you want to give the other person control or allow him to feel that he is in charge of the situation if for example, you were making an apology.
While the palm-up handshake can communicate a submis-sive attitude, there are sometimes other circumstances to consider. As we have seen, a person with arthritis in their hands will be forced to give you a limp handshake because of their condition and this makes it easy to turn their palm into the submissive position. People who use their hands in their profession, such as surgeons, artists and musicians, may also give a limp handshake, purely to protect their hands. The gesture clusters they use following their handshake will give further clues for your assessment of them - a submissive person will use more submissive gestures and a dominant person will use more assertive gestures.
How to Create Equality
When two dominant people shake hands, a symbolic power struggle takes place as each person attempts to turn the other's palm into the submissive position. The result is a vice-like handshake with both palms remaining in the vertical position and this creates a feeling of equality and mutual respect because neither is prepared to give in to the other.
4 Reasons Online Dating Works for Expatriate Singles in Germany
No matter who you ask, you will get the same answer: dating nowadays is hard. For single expats in Germany, dating is even harder. Online Dating
10 Tips for Creating the Perfect Online Dating Profile for Expats
In a perfect world, you and your soulmate would bump into each other on the streets of Germany, lock eyes, and fall madly in love the next second. Dating Profile
Online Dating Tips for Men vs. Women
Is online dating easier for single female expats in Germany than for their male counterparts? Dating Tips
The Research Whisperer
Dr Wade Kellyis the Executive Advisor, Research Impact, at La Trobe University, in Melbourne, Australia.
Wade’s PhD research focused on how and why universities and academics engage with communities.
Image from Tim J. | unsplash.com
Multiple times a year I provide impact statement workshops. Not everyone can make those, so rather than having that knowledge only live in the workshop space, I thought I’d highlight some of the main take-aways shared during that workshop here.
While I’m based in Australia and tailor a lot of my advice to Australian frameworks, this post has relevance for any researcher who is trying to write a research impact statement. If you’re not sure what is meant by research impact, my earlier post about the emerging impact landscape may be helpful.
Increasingly, impact statements are integrated into funding applications. The impact statement should be a standalone snapshot of your project. Impact statements may go by other names — for example, the Australian Research Council’s National Interest Test could be considered an impact statement — but they are essentially asking researchers to provide the same components.
Impact statements are a genre of writing and, once you have a handle on the components that make up a persuasive impact statement, they’re much less daunting to generate.
Types of impact statements
Retrospective — the impact, and how it was generated, from a previous project
Prospective — what you’re planning to do to generate impact for a proposed project
It’s important that you know what your funder is asking for. Some funders want to ensure that you have a firm plan in place to generate impact for a proposed project. Others want proof that you know how to generate impact because you have in the past.
Both retrospective and prospective impact statements have the same components the tense is just different (e.g., “we did” vs. “we will”). To simplify things a little, we’ll stick with prospective impact for statements for the rest of the post.
Impact Statement Components
At the simplest level, there are five components that make a strong impact statement: problem, solution, beneficiaries, outputs, and impact.
The problem statement needs to be high-level, clear, and simply stated. This is the ‘why’ of the research, not just the ‘what’. Problem statements can be stated in the negative (e.g. “childhood obesity is an epidemic”) or as a positive (e.g. “our project will reduce childhood obesity”). In the latter, the reviewer fills in the gap and goes, “Ah yeah, this is a project about childhood obesity.” If you’re having difficulties getting started, ask yourself, “Why is my research needed?”. Pretend you’re explaining it to someone you meet around a campfire or at a barbeque.
This section tends to pose the greatest challenge to researchers when crafting a persuasive impact statement. My suspicion is that researchers know too much about their topic and have difficulty articulating succinctly what the problem is for a generalist audience. Remember, expert reviewers are experts in their subject, not always yours.
In this section, you talk about your proposed solution to the problem. This is where you discuss the research you want to do. Remember to align your proposed solution to the problem statement. There should be a cause and effect relationship here — because this thing is an issue, we’re going to do this. Some grants have multipage impact statements but usually it’s a fairly small space so keep it short, succinct, and high-level.
An issue I often see with this section is that researchers rehash what they have written in other sections of the application. Don’t copy and paste. Start fresh and try not to get too bogged down in the details. You might mention that you will use interviews, but you don’t need to delve into sampling criteria, which generally belongs elsewhere.
Reviewers want to know that you have a clear sense of who is going to be interested in using your research findings. The more concrete your beneficiaries can be, the better. If you can name project partners, name them. If you can say how they will be involved in the project, do.
Saying that your research will benefit the public won’t cut it. It’s just too vague. Indicating that your work, for example, “will benefit young males between the ages of 18-25 by working with x, y, and z organisations,” is much more persuasive.
Outputs could be anything a website, educational resource, videos, pamphlets, workshops, etc. The outputs are how you will get from research results to impact. Impact doesn’t just magically happen so lead the reader on a journey of how you’ll work with beneficiaries and stakeholders to make it happen.
This is another area that researchers seem to struggle with or haven’t though through fully. I suspect it’s because outputs happen after research results so it’s sometimes difficult to conceptualise what forms the outputs may take. Be that as it may, reviewers want to know that you’ve thought through how you will mobilise results into practice.
Often, the stated impact is only a sentence or two. It’s what all your activities will ideally culminate in. It could be an impact on policy, culture, health/wellness, the environment, and/or the economy. If your research is adopted into practice, what will that look like? What will the impact be?
Again, compelling specificity is the key. Lines like, “the impact will be on policy by working with policymakers” are anaemic and won’t convince reviewers that you have a clear sense of how to generate impact. What policymakers are you going to work with? What policy are you looking to impact? There are many posts about impact and policy, including Analysts, advocates and applicators, From science to action, and The hard labour of connecting research to policy during COVID-19.
Putting it Together
In my workshops, I ask people to explain the elements of their impact statements and have others paraphrase it back. Often, when people hear the ways people interpret what they said, they find gaps in logic and issues with clarity.
You can do this exercise yourself as well. Pick up your phone, go to the memo app, and press record. Talk through, to yourself, each of the above impact statement components. Then go for a walk and listen to yourself. Listen a few times and you’ll start to hear where there are jumps in logic, where you have used jargon, where a reviewer might raise their eyebrows.
Write each of the sections above separately — one paragraph each — then stitch it together.
For my workshop, I invite participants to colour-code the impact statement components for their research.
Here’s a marked up example of text from the engagement and impact submission of this study about reforming youth mental health interventions:
You want to ensure that there’s a reasonable balance between each of the sections.
If 90% of the statement is the problem, or the solution, or the outputs, it’s out of balance. You need to make sure that each component is adequately addressed.
Head of the
Department of Conservation Biology & Social-Ecological Systems
Helmholtz-Centre for Environmental Research - UFZ
Research and related interests
• Ecology of Insects in Cultural Landscapes
• Biodiversity, Land Use, and Social-Ecological Systems
• Inter- and Transdisciplinary Cooperation and Projects
• Science-Society-Policy Interface
Sachverständigenrat für Umweltfragen (environmental expert panel of the German government)
Review Board of DFG (German Research Foundation) for the field of “Agriculture, forest science and animal medicine: Ecology of agricultural landscapes” (Fachkollegiat) elected by the German scientific community.
IPBES (Intergovernmental Platform of Biodiversity and Ecosystem Services) Assessment on pollinators and pollination: Drafting co-author of SPM Coordinating lead author (CLA) of chapter 02 “Drivers of pollination and pollinator change“
Prof. h.c., Babes-Bolyai University, Cluj-Napoca, Romania
Review Board of DFG (German Research Foundation) for the field of “Zoology: Biodiversity and Ecosystems” (Fachkollegiat) elected by the German scientific community.
Professor of Ecology (apl.), Martin-Luther-University of Halle-Wittenberg, Germany
Scientific Coordinator ERA-Net BiodivERsA und FACCE-JPI-Projekt STACCATO („SusTaining AgriCultural ChAnge Through ecological engineering and Optimal use of natural resources“)
IPBES (Intergovernmental Platform of Biodiversity and Ecosystem Services) Asia-Pacific Regional Assessment - Lead author (LA) of chapter 04 “Drivers“
IPBES (Intergovernmental Platform of Biodiversity and Ecosystem Services) Assessment on pollinators and pollination - Co-editor of the entire report Drafting co-author of SPM Coordinating lead author (CLA) of chapter 02 “Drivers of pollination and pollinator change“
Scientific Coordinator BMBF Project LEGATO („Land-use intensity and Ecological Engineering – Assessment Tools for risks and Opportunities in irrigated rice based production systems“) ca. 100 scientists and stakeholders from Europa and South-East-Asia, esp. Germany, Philippines and Vietnam
IPCC AR5 Working Group II Coordinating lead author (CLA) of chapter 04 “Terrestrial and inland water systems”
Scientific Coordinator FP6 ERA-Net BiodivERsA-Project CLIMIT („CLimate change impacts on Insects and their MITigation“) ca. 70 scientists of 20 partner organisations from ca. 10 countries CLIMIT
Deputy co-ordinator of the EC FP6 Policy support project MACIS "Policy options to minimise negative effects of climate change on biodiversity"
Head and scientific co-ordinator of the EC FP6 Integrated Project ALARM ("Assessment of LArge scale environmental Risks for biodiversity with tested Methods") ca. 250 scientists of 68 partner organisations from 35 countries
"Privatdozent" 1 of 'Ecology' at the Martin-Luther-University Halle-Wittenberg
1 Privatdozent: a lecturer at German universities who is not a member of the salaried university staff, but is eligible for a professorship
Head and scientific co-ordinator of the EC FP5 RTD Project MACMAN ("Maculinea Butterflies of the Habitats Directive and European Red List as Indicators and Tools for Habitat Conservation and Management") ca. 100 scientists of ca. 40 partner organisations from 18 countries
Deputy Head of Department of Community Ecology at the Helmholtz Centre for Environmental Research UFZ. Head of Working Group on “Animal Ecology and Social-Ecological Systems”
"Privatdozent" 1 of 'Landscape Ecology and Agroecology' at the University of Hohenheim [Habilitation thesis (translated): 'A grid based analysis of metapopulations']
1 Privatdozent: a lecturer at German universities who is not a member of the salaried university staff, but is eligible for a professorship
Deputy Head of Department of Conservation Biology and Natural Resources at the Helmholtz Centre for Environmental Research UFZ
Scientific co-ordinator of and researcher within large interdisciplinary conservation biological research projects (e.g. FIFB = Species survival in fragmented landscapes). Focus on integration of nature conservation and agriculture
Ph.D. degree in agricultural sciences thesis: "Effects of intensification in irrigated rice production on terrestrial arthropods of Philippine Rice Terraces"
Civil service in a natural history museum (Bad Dürkheim, Germany) activities in nature conservation, entomofaunistics and curation of entomological collections)
Scientific assistant at the Agroecology Section of the Institute of Plant Production in the Tropics and Subtropics at Hohenheim University
Second long-term stay in the Philippines for continuation of field studies in the rice terraces of Ifugao Province (basic data collection for Ph.D. DAAD scholarship)
Graduate studies in agricultural biology at Hohenheim University master degree in agricultural biology (thesis: effects of intensification of the rice terrace system on terrestrial arthropod communities) majors in: plant protection, tropical and subtropical plant production, plant ecology and landscape planning
First long-term stay in the Philippines for field studies in the rice terraces of Ifugao Province (basic data collection for master thesis scholarship of private foundation: "Vater-und-Sohn-Eiselen-Stiftung, Ulm, Germany")
Undergraduate studies in agricultural biology at Hohenheim University (agricultural university near Stuttgart, Southern Germany)
Watch the video: Επιπτώσεις από τη ρύπανση των θαλασσών (May 2022).