This is a study of factors which may influence responses to leading questions within a questionnaire format. As such, this study provides insight into how and why leading questions exert social influence in the expression of attitudes. Following an extensive interdisciplinary review of the literature, the effects of questionnaire wording are initially proposed as a form of experimenter bias; it is then pointed out that the respondents' motives in participating in a study have been identified as key to understanding the experimenter bias process. From there, the motives associated informational social influence, and normative social influence, are proposed as intervening motives when subjects yield to leading questions. The personality variables of social desirability (or approval dependency) and self-monitoring were predicted to predispose individuals to yield to leading questions. This investigation was conducted within an opinion poll paradigm, whereby subjects completed an opinion poll with leading questions embedded within it; the leading questions for this study suggest their desired responses through parenthesized, numeric suggestions, or cognitive anchors. In order to test the various hypotheses, different conditions were created by varying the introductory remarks of the opinion poll. Support for the role of informational social influence in yielding to leading questions was provided by the finding that people yield more to leading questions which are apparently designed by experts (professional pollsters) than by nonexperts (highschool students). Support for the role of normative social influence was provided by the significant association between responses which subjects provide for themselves, and responses which they provide to represent those of their peers--i.e., subjects they tend to assume that their biased responses are congruent with those of their peers. The personality variables of social desirability and self-monitoring were also found to affect responses to leading questions; the relationship between these two variables appears to be interactive, and not compounded as predicted. A post-hoc study suggests that the results may be confounded by the situational demands of the questionnaire, such as a situational demand for "honesty". This study's particular form of leading questions may have also confounded the results, and more varied forms of leading questions would be worthy of future investigation.
| Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/6771 |
| Date | January 1993 |
| Creators | Milstone, Carol. |
| Contributors | Edwards, Henry P., |
| Publisher | University of Ottawa (Canada) |
| Source Sets | Université d’Ottawa |
| Detected Language | English |
| Type | Thesis |
| Format | 166 p. |
Page generated in 0.0014 seconds