Case Study

Based on your comprehension of the methodology used by Jouriles and colleagues, explain why the researchers decided to dichotomize their measures of aggression rather than using the raw skewed frequency scores, and explain how this transformation changes what is being compared.

Case context: A research team is studying relationship aggression among high school students. They gather survey data on the frequency of physical aggression incidents. Upon reviewing the raw data, they find a heavily right-skewed distribution, where the vast majority of participants report 00 incidents, and only a tiny minority report one or more incidents. The team is worried that using these highly skewed distributions will violate statistical assumptions when computing correlations.

Question: Based on your comprehension of the methodology used by Jouriles and colleagues, explain why the researchers decided to dichotomize their measures of aggression rather than using the raw skewed frequency scores, and explain how this transformation changes what is being compared.

Sample answer: Jouriles and colleagues dichotomized the aggression measures because measures of physical aggression often result in highly skewed distributions that violate assumptions of normality required for standard correlation analyses. By converting the scores to binary values (00 for did not occur and 11 for did occur), the researchers transformed the continuous frequency counts into simple categories of presence versus absence of aggression. This allows them to examine whether the occurrence of one type of aggression is correlated with the occurrence of another, bypassing the issues caused by extreme skewness.

Key points:

  • Aggression measures often yield highly skewed distributions because most adolescents report zero incidents.
  • Dichotomizing converts the skewed distributions into binary variables (00 and 11).
  • The binary variables represent the simple occurrence (11) vs. non-occurrence (00) of aggression.
  • This transformation allows researchers to compute correlations among the constructs without the distortion caused by highly skewed raw data.

Rubric: The explanation must correctly identify that dichotomizing addresses the issue of highly skewed distributions that violate statistical assumptions. It must also explain that the transformation changes the analysis from assessing raw frequency of events to assessing the simple occurrence vs. non-occurrence of those events.

0

1

Updated 2026-05-27

Contributors are:

Who are from:

Tags

KPU

Research Methods in Psychology - 4th American Edition @ KPU

Related