H567/02 Section A - Bocchiaro et al. (2012): disobedience and whistle-blowing

Three responses to authority

Bocchiaro, Zimbardo and Van Lange extended obedience research by giving participants an opportunity to report an unjust authority.

Milgram showed harmful compliance under pressure, but refusal alone does not expose a wider system. The contemporary study asked how people respond to a researcher requesting deceptive recruitment for an apparently unethical sensory-deprivation study. It also examined whether predictions, personality and values distinguish those who obey, refuse or report misconduct.

In the study, writing the requested endorsement without reporting counted as obedience. Refusing without reporting counted as disobedience. Reporting to the Research Committee counted as whistle-blowing, whether the participant had written the endorsement or refused. Thus writing an endorsement and whistle-blowing could occur in the same person; the final three groups were classified with reporting taking priority.

This is social psychology because an authority's request, institutional legitimacy and responsibility towards fellow students shape behaviour. Its theme is Responses to people in authority, including active challenge as well as compliance.

Sample and measurement

The final behavioural sample consisted of 149 VU University Amsterdam undergraduates recruited through campus-cafeteria flyers.

There were 96 women and 53 men, mean age 20.8 years, with standard deviation 2.65. Students received €7 or course credit. Eleven of the initial 160 were excluded for suspicion about the study. This self-selected, student sample includes both genders but is not a random sample of Dutch adults.

The authors describe an experimental scenario. All main participants received the same unethical request, so there was no manipulated independent variable separating experimental groups. A controlled laboratory scenario study is a precise description; group membership was the measured response, not random assignment to obedience or whistle-blowing. Personality–response associations are correlational.

After the behavioural task, the Dutch 60-item HEXACO-PI-R measured honesty–humility, emotionality, extraversion, agreeableness, conscientiousness and openness. A nine-item Decomposed Games measure assessed social value orientation: preferences for joint/equal outcomes, personal gain or relative advantage. The study combined observable acts with numerical self-report scores, giving different kinds of evidence rather than one interchangeable measure of morality.

A separate 138-student comparison sample read the scenario and predicted their own behaviour and that of an average student. They did not undergo the request; differences from the main sample are therefore differences between prediction and observed action, not before-and-after changes in the same individuals.

Unethical recruitment procedure

A standardised cover story created a conflict between obeying a researcher and protecting other students.

Eight pilots involving 92 undergraduates tested credibility, acceptability and consistent authority behaviour. In the main study a formally dressed, stern male Dutch experimenter met each student individually, explained participation rights and confidentiality, and requested names of fellow students. He then described a supposed earlier sensory-deprivation study in Rome: participants panicked, some hallucinated, and requests to stop had been refused. He said that another study was planned at VU and that its Research Committee was considering approval.

The participant was asked to persuade the named students to volunteer. After three minutes alone to reflect, the participant moved to a second room containing a computer, committee forms and a mailbox. The recruitment message had to use at least two enthusiastic adjectives from a specified set and omit the negative effects. The experimenter left for seven minutes, providing an opportunity to comply, refuse and/or anonymously report the ethical problem.

The experimenter returned, administered the personality/value measures, checked suspicion and carefully debriefed the participant. The full session lasted about 40 minutes. Debriefing explained the deception, sought to reduce shame and distrust, supplied written information and a contact address, and obtained informed consent for the use of the data. The proposed harmful deprivation research was a cover story; participants did not actually endure it.