arnold on Sep 16, 2008
Schriescheim et al (1989) wrote an an article where they conducted three studies exploring the effect of two questionnaire formats (grouping versus randomized items) on internal consistency and test-retest reliability, convergent and discriminant validity, and accuracy in respondent-provided descriptions.
Conclusion: They found that neither format (group/random) is superior to the other on any of the examined psychometric properties, although a weak trend showed that the grouped item format produced slightly superior measurement properties, but did not support the usefulness of grouping.
Schriesheim, Kopelman, Solomon (1989). “The Effect of Grouped Versus Randomized Questionnaire Format on Scale Reliability and Validity: A Three-Study Investigation.” Educational and Psychological Measurement, 49, 487 – 508.