Ranking statements against each other removes the easy answer, which is the main structural defence against deliberate distortion.
Why it matters when the plan changes
High-stakes assessment gives respondents a clear incentive to manage their answers, and rating scales make it easy: each item is scored independently, so a respondent can simply endorse every desirable statement, which is the mechanism behind social-desirability bias, the tendency to answer in ways that will be viewed favourably. The people most motivated to present well are often the people being considered for the most consequential roles, so the distortion concentrates exactly where accuracy matters most.
The tension is that no instrument is faking-proof, and claiming otherwise is itself a warning sign. Format makes distortion harder because forced-choice ranks statements against each other rather than scoring them independently, removing the single favourable answer a rating scale allows; detection catches some of what still gets through. A supplier who describes their instrument as impossible to fake has overclaimed, and the honest framing is always reduction of the problem rather than its elimination, which is what a buyer should expect to be told.
In practice
Two candidates complete an assessment. One answers honestly and one answers as the person they think the role wants. On a rating scale both produce strong, similar profiles. In a forced-choice format the second candidate has to choose between equally attractive statements and the pattern of those choices carries information their self-presentation cannot control.
Evidence
Respondents tend to answer in ways that will be viewed favourably, over-reporting good behaviour and under-reporting bad.
Social-desirability bias, Wikipedia (2026)Forced-choice scales ask respondents to rank statements against each other, contrasting with rating scales where each item is scored independently.
Ipsative, Wikipedia (2026)
What it cannot tell you
Faking resistance cannot tell you whether an instrument is unfakeable, only that distortion is harder and some of it is caught. It says nothing about the accuracy of the underlying trait measure itself, and detection flags indicate unusual response patterns, not proof of intent to deceive.
Questions
To some degree, yes. Social-desirability bias, documented in the Wikipedia entry on Social-desirability bias (2026), shows respondents tend to answer favourably regardless of format. Forced-choice and detection reduce how easy and effective distortion is; neither eliminates it, and a supplier claiming an unfakeable instrument has overclaimed.
Forced-choice scales, sometimes called ipsative scales, ask respondents to rank statements against each other rather than scoring each independently, as documented in the Wikipedia entry on Ipsative (2026). This removes the option of agreeing with everything desirable, so a respondent's choice between two attractive statements carries information their self-presentation cannot control.
They flag response patterns that are statistically unusual for honest answering: unlikely combinations, excessive consistency, or answers that are too uniformly favourable. A flag is a reason to treat a result cautiously or re-administer, not a conclusion that someone was dishonest.
It can, because people differ in how much they distort. If everyone inflated equally the ranking would survive; they do not, so distortion reorders candidates. That reordering is invisible in the scores, which is why format matters more than any warning to respondents.
Being transparent that response patterns are checked is honest and tends to reduce distortion slightly. What should not happen is using a detection flag as an accusation. It is a signal about data quality, and the appropriate response is usually another conversation rather than a conclusion.