The CFA also notes evidence of method variance (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003) beyond the hazard factors in the large error correlations for four sequential items; these correlations indicate a response set across similar items within a questionnaire. Variations in the ordering of items may help to reduce the impact of method variance.