THE EFFECT OF NEGATION AND POLAR OPPOSITE ITEM REVERSALS ON QUESTIONNAIRE RELIABILITY AND VALIDITY - AN EXPERIMENTAL INVESTIGATION

This research was concerned with the effect of item wording on questionnaire reliability and validity. The subjects were 280 business undergraduate students who read a scenario describing a hypothetical leader's behavior and then completed one of four different questionnaires to describe that leader. Comprising the experimental treatments, the questionnaires included four different types of items: regular (e.g., "I am happy"), polar opposite (e.g., "I am sad"), negated polar opposite (e.g., "I am not sad"), and negated regular (e.g., "I am not happy") items. The results indicated significantly lower coefficient alpha internal-consistency reliabilities for the polar opposite and negated polar opposite items as compared with those for the regular and negated regular items. Using accuracy scores, one-way analyses-of-variance and post-hoc directional t-tests showed the same results for validity of measurement. Implications for future questionnaire development and research are briefly considered.