Can you give specific examples of single behaviors which have that effect? Or, if single behaviours do not reliably have that effect, a minimal set of distinct behaviors that do have that effect?
There probably aren't going to be any, due to what the PUAs call "shit testing".
The term refers to deliberate probing behaviors intended to determine someone's status (and therefore attractiveness), by finding out how confident they are in response to resistance, criticism, etc.
This potentially lets the tester get past any faked confidence/status signals, and appears to be triggered by any incongruity or sense that the status-displaying person is "too good to be true".
Or in more Bayesian terms, when an approaching man's status displays indicate higher status than a woman's prior for the status of men approaching her, she will be more inclined to respond in ways that "test" him, in order to determine his true status.
However, the PUAs also generally contend that this process never stops, although it occurs more at first meeting and at any perceived status increase, with "maintenance" testing occurring throughout a relationship.
During these testing periods, however, testing behaviors usually escalate, and rely heavily on immediate, unconscious reactions rather than on consciously controlled behaviors... which are much harder to fake.
I recall one experience I had -- not trying to pick anyone up, but just talking with a woman who was displaying attraction signals after I sociably teased and picked on her a bit. At one point, she asked me something that I declined to do, and I made some sort of status-implying joke about it.
She then asked me in an almost-angry tone, "Are you for real?" I hesitated, for just a fraction of a second... and then the dynamic changed. Her previous displays of attraction disappeared in an instant.
At the time, I reacted to the words and the tone, but when I replayed in my mind what I saw at the time, I could see that she really did want to know if I was "for real", was hoping I was (very expectant look, slight hopeful upturn of tone at the end of the sentence), and was then disappointed (slumped shoulders, vacant gaze) to see my hesitation... and therefore, my non-realness.
Interestingly, the point at which this occurred was precisely the point at which I had attempted a status display that I did not truly feel; i.e., one that was higher than my own perception of my status in the interaction. All of the status-implying jokes I made before that point were ones I was 100% confident of, whereas that one I wasn't entirely sure of.
On a more recent note, my wife tends to make "testing" comments or actions whenever I make mental changes that positively affect my interactions with her... and she knows that she does it, but doesn't know how to stop. She describes it as an impulsive feeling that I'm being "too good to be true" and of wanting to know if I'm "for real", with things blurting out of her mouth before she can stop them.
Both of these experiences seem to support the idea that status testing is an interactive process, at least in the context of men's attractiveness to women. Which is probably why some PUA gurus emphasize practice so much (so that their students' status moves will be natural), while others emphasize working on personal centeredness (so that their students won't appear hesitant or unsure of themselves, no matter what other people do or say).
IOW, the overt status signals being discussed here are only bids for status, in the attractiveness context. Actual status is determined by whether you can back up your bid, or more precisely, by whether you appear 100% confident in your ability to back up your bid, as measured by your response to challenges.
The "status" hypothesis simply claims that we associate one another with a one-dimensional quantity: the perceived degree to which others' behavior can affect our well-being. And each of us behaves toward our peers according to our internally represented status mapping.
Imagine that, within your group, you're in a position where everyone wants to please you and no one can afford to challenge you. What does this mean for your behavior? It means you get to act selfish -- focusing on what makes you most pleased, and becoming less sensitive to lower-grade pleasure stimuli.
Now let's say you meet an outsider. They want to estimate your status, because it's a useful and efficient value to remember. And when they see you acting selfishly in front of others in your group, they will infer the lopsided balance of power.
In your own life, when you interact with someone who could affect your well-being, you do your best to act in a way that is valuable to them, hoping they will be motivated to reciprocate. The thing is, if an observer witnesses your unselfish behavior, it's a telltale sign of your lower status. And this scenario is so general, and so common, that most people learn to be very observant of others' deviations from selfishness.
On Less Wrong, we already understand the phenomenon of status signaling -- the causal link from status to behavior, and the inferential link from behavior to status. If we also recognize the role of selfishness as a reliable status signal, we can gain a lot of predictive power about which specific behavioral mannerisms are high- and low-status.
Are each of the following high- or low-status?
1. Standing up straight
2. Saying what's on your mind, without thinking it through
3. Making an effort to have a pleasant conversation
4. Wearing the most comfortable possible clothes
5. Apologizing to someone you've wronged
6. Blowing your nose in front of people
7. Asking for permission
8. Showing off
Answers:
1. Standing up straight is low-status, because you're obviously doing it to make an impression on others -- there's no first-order benefit to yourself.
2. Saying what's on your mind is high-status, because you're doing something pleasurable. This signal is most reliable when what you say doesn't have any intellectual merit.
3. Making an effort to have a pleasant conversation is low-status. It's high-status to talk about what you care about.
4. Wearing the most comfortable possible clothes is high-status, because you're clearly benefiting yourself. (Dressing in fashionable clothes is also high-status, through a different inferential pathway.)
5. Apologizing is low-status because you're obviously not doing it for yourself.
6. Blowing your nose is high-status because it's pleasurable and shows that you aren't affected enough by others to stop.
7. Asking for permission is low-status. Compare: recognizing that proceeding would be pleasurable, and believing that you are immune to any negative consequences.
8. Showing off is low-status, because it reveals that the prospect of impressing your peers drives you to do things which aren't first-order selfish. (Of course, the thing you are showing off might legitimately signal status.)
Pwno's post makes a good related point: The most reliable high-status signal is indifference. If you're indifferent to a person, it means their behavior doesn't even factor into your expectation of well-being. It means your computational resources are too limited to allocate them their own variable, since its value matters so little. How could you act indifferent if you weren't high-status?