Self-report surveys can manufacture false correlations when one person supplies all data.

Evidence

Common Method Variance Another potential problem endemic to survey data is common method variance, which occurs when the predictor data is collected from the same source as the outcome data (MacKenzie, Podsakoff, & Rich, 2001; Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). The relationships reported may simply be a result of the same respondents providing the responses for both sets of variables. — Roderick M. Kramer, Geoffrey J. Leonardelli, Robert W. Livingston, Social Cognition, Social Identity, and…, 1861

Categorization

• QUESTION: Worth noting as a standing methodological caveat when evaluating any survey-based claims about prejudice or identity—how often does this taint the psychology literature I’m reading?
• corroborated: 2 independent passages support this claim — Roderick M. Kramer, Geoffrey J. Leonardelli, Robert W. Livingston, Social Cognition, Social Identity, and…, 1861