A procedure for isolating social desirability variance in a three-way component analysis

Area of Science:

• Psychometrics
• Personality Psychology
• Data Analysis

Background:

• Three-way component analysis summarizes complex data from individuals, items, and situations.
• Social desirability response bias can distort findings in questionnaire data.
• Existing methods may not adequately address this bias in three-way analysis.

Purpose of the Study:

• To propose and evaluate a preprocessing procedure for controlling social desirability bias.
• To isolate and remove social desirability variance from three-way data sets.
• To improve the accuracy of component analysis in personality research.

Main Methods:

• Developed a novel preprocessing procedure to identify and remove social desirability variance.
• Applied the procedure to empirical data from the personality domain.
• Conducted a simulation study to assess the method's performance and impact.

Main Results:

• The proposed method effectively controlled social desirability bias in the data.
• Bias led to increased person variance and decreased triple interaction effects (persons × items × situations).
• Bias also caused partial distortion of the person component and core matrix, which was mitigated by the procedure.

Conclusions:

• The preprocessing method is effective in controlling social desirability bias in three-way questionnaire data.
• Removing this bias enhances the reliability and validity of component analysis results.
• This approach offers a valuable tool for researchers in personality and related fields.