Related Experiment Videos
How do respondents construe ambiguous response formats of affect items?
Susana Lloret Segura1, Vicente Gonzalez-Roma
1Department of Methodology of Behavioral Sciences, Faculty of Psychology, University of Valencia, Valencia, Spain. susana.lloret@uv.es
Journal of Personality and Social Psychology
|November 6, 2003
Summary
Researchers developed a method to understand how people interpret survey response formats. The study found that respondents consistently interpreted opposite items as bipolar, not unipolar.
Area of Science:
- Psychometrics
- Quantitative Psychology
- Survey Methodology
Background:
- Ambiguous response formats in surveys can lead to misinterpretations.
- Item response correlations can range from -.467 to -1, depending on unipolar or bipolar interpretations.
- Distinguishing between unipolar and bipolar interpretations is crucial for accurate data.
Purpose of the Study:
- To present a novel procedure for investigating the interpretation of response formats for opposite items.
- To determine whether respondents interpret item pairs as unipolar or bipolar.
- To identify the underlying response format construction.
Main Methods:
- Application of nonparametric (Mokken model) and parametric (graded response model) item response theory (IRT) models.
- Analysis of item pairs with opposite adjectives (e.g., "relaxed"-"tense").
- Testing the procedure across four independent samples.
Main Results:
- The proposed procedure successfully identified the response format construction.
- Respondents consistently interpreted the response formats as bipolar.
- The results provided empirical support for the bipolarity of the tested item pairs.
Conclusions:
- The developed procedure is effective in discerning response format interpretation.
- Survey respondents tend to use bipolar interpretations for opposite items.
- Findings have implications for improving survey design and data analysis in psychometric research.