Faking "turn-ons" and "turn-offs". The effects of competitory covert imagery on penile tumescence responses to diverse extrinsic sexual stimulus materials.

One adult man served as his own control in a lab. Researchers showed him sexual films while they recorded penile size with a strain gauge.

They also told him to use private thoughts: either think sexy thoughts during neutral clips or think boring thoughts during sexy clips. They wanted to see if he could fake the results.

The man could shrink his response to sexy films by picturing dull things. He could also grow his response to neutral films by picturing sexy things.

His body fooled the machine both ways. The readings no longer matched what was on the screen.

DeLeon et al. (2005) found the same risk in preference tests. Fixed five-minute trials missed the best toys for reducing problem behavior. Both papers warn that standard timing can hide true results.

Falligant et al. (2020) show another pitfall. Dual-criteria visual aids can give false positives if Phase B is too short. Together these three papers say single-case data need extra checks, no matter the measure.

Richman et al. (2001) and Tavassoli et al. (2012) looked at theory-of-mind tasks. They asked whether kids can fake answers or if the test itself misses skill. The theme is the same: if clients can game the task, or if the task is blind, your data are weak.

If you use penile tumescence, eye-tracking, or any indirect measure, add validity steps. Ask the client to describe thoughts during the test. Run brief catch trials with known stimuli. Compare results across two sessions. One lab subject beat the machine; your client might too.