Research Methods II: Social Psychology Issues

Research Methods II: Introduction to Social Psychology

• This presentation covers key issues in social psychology research.

The Replication Crisis

• The current methods of research and reporting findings are leading to a higher rate of false positives.
• This is a significant contributor to replication failures.
• Common issues include:
  • p-hacking
  • Cherry-picking data
  • HARKing (Hypothesizing After Results are Known)
  • Other questionable research practices

Key Research in Social Psychology

• Stanford Prison Experiment (Philip Zimbardo):

  • Examined the psychological effects of perceived power.
  • Questioned whether brutality among prison guards is due to sadistic personalities or a situational effect (prison setting).
  • Ordinary students assigned as guards showed unexpectedly brutal behavior.
  • This experiment's methodology and conclusions have been questioned, mainly referring to the highly directed nature and a lack of validity.

• Robber's Cave Experiment (Muzafer Sherif):

  • Investigated "Realistic Conflict Theory".
  • Study involved 11-13 year-old boys in a summer camp, assigned to two groups ("houses").
  • A competition created conflict and hostility between the groups.
  • The ultimate resolution involved a common goal (e.g., a common objective) that brought the groups back together.
  • The findings have been questioned, mentioning factors like how the study's conditions were artificially set up and factors of researcher influence.

• Implicit Social Priming Experiments:

  • Explore "Automaticity Effects".
  • Use subtle prompts (words, images, experiences) to activate concepts in the mind of participants.
  • Observe participant behavior and its alignment with activated concepts.
  • This could reveal the influence of subconscious cues on social interactions.
  • However, experimenter biases are a concern in this type of research which raise question over the validity of the study.

Problems in Social Psychological Research

• Experimenter effects and biases: Experimenters are aware of the experiment conditions so this could unintentionally influence results. They could sometimes change their behaviour on purpose or through accident; these changes impact participants' behaviours and this could lead to conclusions that don't reflect what happened.
• Ethical concerns: In regards to whether or not results speak to broader theories or are merely an experimental difference between the variables.

How to Enhance Research Outcomes

• Use Simmons et al's (2012) suggestions: Fully disclose how sample sizes, data exclusions, and other relevant factors were determined.
• Pre-register empirical studies: Outline hypotheses, variables, procedures in advance.
• Post data and results: Share the raw data and results to foster transparency.

Reducing P-hacking

• Understand p-hacking as a data "fishing expedition".
• Use p-curve analysis to identify p-hacking and publication bias.
• Using appropriate "p-value" thresholds to avoid bias during research. p-values around .01 and .02 rather than .04 or .05 is seen as a key indicator of whether p-hacking has taken place or not.

Reducing HARKing

• Hypothesize Before Collecting Data: Formulate hypotheses before collecting the data, rather than afterward.
• Clarity in Hypotheses: State hypotheses explicitly; they should be comprehensive, directional, and parsimonious.

Hypothesis Examples

• Bad vs. good hypotheses and examples of how to generate better hypothesis statements.

Improve Statistical Power

• Focus on within-subject designs.
• Use a larger sample size.
• Understanding Cohen's effect size guidelines (small, medium, and large effects).

Take-Away Points & Good News

• Emphasize replication as being just as important as innovation.
• Encourage the use of methods like registered replication reports, for greater transparency, and a better understanding in the conclusions of research.
• Highlight the recent trends in psychology of adopting open-science practices and using social media for dissemination.