Social Psychology: Replicating Milgram - Burger (2009)

Burger (2009) Replicating Milgram

Aim

To investigate obedience by partial replication of Milgram’s study almost 50 years later to examine whether situational factors affect obedience to an authoritative figure.

Participants

• 29 males and 41 females aged 20-81 years old from a range of backgrounds.

• Volunteer sampling recruited via newspaper advertisements.

• Participants were paid $50$ for participating.

Screening of Participants

Volunteers familiar with Milgram’s work were excluded and screened for mental health and drug dependency using:

• Clinical interviews

• Empathetic Concern scale (Interpersonal Reactivity Index)

• Beck Anxiety Inventory

• Desirability of Control measure

• Beck Depression Inventory

Procedure

• Laboratory-based experiment at Santa Clara University.

• Participants were randomly assigned to two groups.

• The procedure mirrored Milgram's but only went up to $150$V (Milgram went to $450$V).

• A $15$V real shock was administered (Milgram used $45$V).

• Pre-recorded voice feedback after $75$V for standardization.

Conditions

• Baseline: Learner reveals a heart condition at the beginning; at $150$V, the learner shouts, ‘Get me out of here my heart is starting to bother me now’.

• Modelled Refusal: Two confederates were used; one to model refusal at $90$V.

Results

• Base condition: $70\%$ continued to $150$V.

• Model refusal condition: $63.3\%$ continued to $150$V.

• Experiment 5: $82.5\%$ continued above $150$ V

Conclusion

Obedience rates are still similar to Milgram's, with no statistical difference. There were no significant differences in obedience rates based on gender, age, race, education, or personality.

Ethics

• Informed consent was obtained; participants were told it was a study about the effect of punishment on learning.

• Participants were given the right to withdraw verbally and in writing.

• Debriefed immediately after the experiment, shown shocks weren't real and the learner was unharmed.

• Screened to protect from harm

Evaluation

• Reliability: Standardized procedure allows for replication.

• Internal Validity (Limitation): Assumes those who stopped at $150$V would have continued to $450$V, which cannot be certain.

• Generalisability(Strength): Equal ratios of F to M ppts to reduce sampling bias.

• Ecological/External Validity (Weakness): Lab experiment at Santa Clara University.

• Protection from physical harm (Strength): $45$V v $15$V for real shock and screening (anxiety and depression using Beck scale)

• Still used verbal prods (Weakness): ‘you must continue’; ‘punishment and learning’