The use of affective and cognitive decision-making processes during questionnaire completion is examined through protocol analysis. Characteristics of the questionnaire, such as length and extremity of the scale anchor cues, resulted in more affective processing. Respondent-related variables, such as product familiarity, attitude toward survey research, and sex, had little influence on decision making. Finally, affective decision making influenced the outcomes of the questionnaire response analysis by being associated with more within-item variance, less decision time, and lower ratings. The implications of the results to applied and theoretical research are discussed. © 1994 John Wiley & Sons, Inc.