I know it exists and it has a name...
But my google-fu skills are lacking, due to a pretty niche concept that uses a lot of generic keywords.
There is a fallacy / mental model / a tendency for people to portray themselves in a positive light.
I know about "80% of drivers thinking they are better than average" https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect and https://en.wikipedia.org/wiki/Illusory_superiority but it is not exactly what I'm looking for.
CONCRETE EXAMPLE
Have you ever taken / ingested / consumed cannabis?
Answer "YES" to this question would admit doing something illegal, with potentially negative consquences, including the fact that the researcher might be an undercover cop. Therefore, even if the genuine answer is "YES", there is a non-zero probability that the recorded answer is "NO"
To account for this tendency INSERT NAME HERE _____________
I would like to suggest an alternative design of the same question that involves a 2nd researcher
"We are aware of the tendency
INSERT NAME HERE _____________
"We are working towards legalization efforts..."
"This will be treated in the strictest confidence..."
"We know that asking "YES / NO" questions about drug use can make people..."
"Our goal is to obtain more accurate estimate..."
CONCRETE EXAMPLE 2
3rd April 2022 source (behind a paywall)
Orban’s bid for fourth straight term comes down to knife-edge vote
5th April 2022 source (behind a paywall)
Crushing victory gives Viktor Orban scope to tighten grip on Hungary
Same with Trump and other socially unpopular candidates.
More google-fu
"anonymity bias" - loads of interesting research papers, including one specifically about cannabis (link):
Of course the sample size is small but the discrepancy is clearly visible (other research papers confirm it too).
I am aware that anonymity might not be enough to obtain accurate results, that's why the design of the study with the 2nd researcher but I just want to sound a little bit smarter and read more about INSERT NAME HERE _____________
- surely it has a name?
EDIT: if it does not have a name, I'll take the "Robertson Syndrome"
INSERT NAME HERE _____________
bias / fallacy / phenomenon... "Saving face" sounds broadly related but it doesn't nail it, thank you for the comment :) $\endgroup$