I don't know if I can answer your question, but perhaps this will help. Of course there are different types of self-enhancement - you can be deceiving the other person (e.g., exaggerating), or you could be deceiving yourself. In the latter case, one important ingredient is the amount of ambiguity that the situation allows.

For instance, there are some things that are quite narrowly constrained. You could probably deceive yourself a inch (or centimetre) or two when it comes to your height, but there's not much room to maneuver. You certainly couldn't convince yourself that you were Wimbledon champion. But when there's a lot of ambiguity in the situation, it's much easier to self-deceive, and therefore self-enhance. That's why it's much easier to believe you're a good driver; because it's a largely subjective skill that different people can define different ways (David Dunning has some good research on this point in the 90's with Geoff Cohen, for instance.). Even if you get into several car accidents you could still try to blame the other driver, or the lamp post for getting in the way ;)

So when it comes to sexual partners, I think the same thing holds. If it's a question that doesn't admit much ambiguity (let's say there was only one, then it would be hard to misremember it as two), then there might be less self-deception -- although that doesn't exclude deceptive exaggeration. If there is some ambiguity (like someone who has dated many people over the course of many years) then a great deal more self-enhancement could find its way in.

From a research point of view it is important to realize that it is not just how common self-enhancement is that is important, but how strong the effect is in your sample. Common method variance can result from just a few percent people being prone to deception, with effects that are as large as many other reported in the social science literature (I have done some simple simulations on this). So, if you want to predict number of sexual partners from a risk-taking scale, you will get a correlation that is partly due to the real association, partly to systematic response bias, of which self-enhancement is one part.

Returning to the original question, I would say that most people are prone to self-enhancement, its just the degree that differs. This was the conclusion of Uziel (2010), in his review of lie scale research. Its a personality trait, so its always there. My own research confirm that.

As for constraints on effects, I agree with David (excellent description).

I have also encountered massive effects of socially sensitive situations in my research. People tend to fake if they believe it might be important, and they can get away with it.

Uziel, L. (2010). Rethinking Social Desirability scales: From Impression Management to interpersonally oriented self-control. Perspectives on Psychological Science, 5, 243-262.

Thank you all for taking time in a detailed explanation! Have nice weekend!

Sven, I would need a tighter definition than 'men' for your initial example; and would suggest for instance that most married men significantly decrease their admitted number of sexual partners, especially in conversations with their spouses, although this might have more to do with self-preservation rather than being in line with the 'modesty effect' mentioned by Jerry!

I am concerned with what I see in this conversation as an all-or-nothing definition of the self-enhancement. Even in the context of recounting one's past sexual history, we know that what constitutes "having sex" is a fuzzy concept. Even "intercourse" is a fuzzy concept. Beyond the fact that it might enhance one's standing to claim a large number of sexual partners and it might enhance one standing to claim a small number of sexual partners, depending on the audience, even with the same number of reported partners, one might imply a level of sexual prowess or less so; or one might imply there were many one night stands, or that one has a large amount of sexual experience, even though the number of partners is small; etc. Indeed, knowing that "I have had 42 past partners" may be the opposite of self-enhancing in that it suggests you are keeping count.

I would be tempted to say that self-enhancement is active – as a process/consideration – a great deal of the time, while self-enhancing behavior may be tempered by not knowing what behavior will, in fact, enhance one's standing, by concurrent processes such as self-verification, as well as, of course, other constraints on behavior unrelated to how one sees oneself or is seen by others.

So I think in the end the question does not make sense until one operationalizes self-enhancement. Compare the question: How often does one act in order to increase one's health? Is having your yearly exam with your PCP such a behavior? Eating a healthy meal? Eating any meal? Breathing? Drinking water? Having a positive thought? Catching yourself in a negative self-judgment and cutting yourself some slack? Without a clear definition of what one means by, and how one would measure, health enhancing behavior, the question of how often it occurs is meaningless.

Hope this helps.

Ron, you are definitely right. Things are usually more complicated than in a starting question. However, you and Sephan already might point to a suggestion how to solve the problem: Maybe it is sufficient to consider the subjective intention to self-enhance, that is, the intentional deviation from what the subject thinks is to be true.

Peter, thank you for your enquiry! For the example, I would define "men" as the "common" male psychology student at an age about twenty, used in many psychology studies. Concerning the justified objection about the meaning of culture, I suggest the Western culture of the present. But it seems that, with Sephan's link, my example with sexual partners is not appropiate anyway.

Sephan, thank you for your hint and the link! The paper is impressive. And together with your earlier comment it may point to my initial consideration: If it is not gender, what is it that makes some people prone to self-enhancement than others? There are of course social context and situations that increase self-enhancement. (I think with enough social pressure indeed anybody would be "induced" to self-enhance.) However, individual aspects surely play an important role as well.