The researchers, Anne Jansen and colleagues, drew on 53 recruiters (HR professionals) from a range of Swiss companies, and two adult student groups representing applicants (416 Masters students, replicated with 88 vocational apprentices). Both recruiters and applicants were presented with a set of self-presentation behaviours, such as "When applying for the job, I praised the organization" or "When applying for the job, I claimed to have experience that I didn’t actually have".
Recruiters were asked how appropriate the behaviours were, and agreement between their responses was high, strongly sharing expectations for half of the behaviours, and moderate agreement for virtually all the remaining. Collectively, they saw some behaviours, such as describing skills or knowledge, as appropriate and uncontroversial, with others definitely inappropriate, such as fabricating details, and still others, strategic ploys such as de-emphasising negative attributes, fell in between. This shared set of norms is what the research team expected, creating a job selection 'situational script' that recruiters expect to be followed. Did the applicants do so?
Enter the dice. Afraid of being tarred a faker, people are reluctant to admit to self-presentation, even for supposedly confidential, anonymous research. To address this, the applicants gave responses using the randomised response technique, which asked them only to reply truthfully to an item if they rolled a three or greater on a playing die - otherwise, they must respond affirmatively, regardless of the truth. This makes individual profiles impossible to identify whilst the aggregate data remains analysable, by looking at how responses differ from the base rate.
Jansen's team examined this data using correlation to compare frequency of applicant behaviour to recruiter judgement of that behaviour; they found high correlations at well above .8 (.9 in the larger Masters sample). The frequency of a self-presentation behaviour was strongly related to whether it was something that recruiters saw as acceptable.
The authors see this as the inevitable outcome of a 'strong situation', with right or wrong ways to behave - the shared attitude of the recruiters - where applicants are just trying to follow that script and do what they are 'supposed to', as learned from advice, previous experience, websites, or tacit feedback from the recruiter. Jansen and her colleagues conclude that common reactions to self-presentation behaviours, such as moral condemnation or celebration as a social skill (not dissimilar to the concept of 'ability to identify criteria'), may be attempts to conjure individual qualities from what is mainly a situational phenomena. Conversely, it seems to me that, as understanding an individual's qualities is so useful in job selection, we would do well to experiment with meeting candidates in weaker, ambiguous situations with no right way to behave, to let them slide off-script and see the real them.