Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.418026
Title: Context effects in social surveys : a study of question order effects
Author: Jonsdottir, Gudbjorg Andrea
Awarding Body: London School of Economics and Political Science
Current Institution: London School of Economics and Political Science (University of London)
Date of Award: 2004
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Abstract:
This research investigates context effects in social surveys. Context effects, the ways in which previous questions affect responses to later questions, can occur when questions deal with aspects of closely related issues. Such effects have been explained by reference to social cognition, social judgement and cognitive consistency theories. Following a review of the literature, three 'split-ballot' experiments, exploring how increasing the salience of particular topics affects responses to related questions, were conducted in a multi-wave national panel survey in Iceland. The first experiment investigates the effects of knowledge questions on expressed political interest. Respondents who were asked the knowledge questions first reported less interest in politics than did other respondents. Contrary to self-perception theory, it is argued that the knowledge questions serve as a frame of reference for the interpretation of the vague phrase, 'to follow what is going on in government and public affairs'. In the second experiment the accessibility of particular topics was increased by providing specific or general examples in the preamble to a question asking about the most important problems facing the nation. The results lend support to the principle of cognitive accessibility, which suggests that respondents do not systematically retrieve all relevant information stored in memory but base their answers on the most accessible information. It is argued that the specific examples widen the question's frame of reference and thus change its meaning. In the third experiment general attitude questions about abortion and nuclear weapons were placed either in pro- or anti-contexts. Respondents who agreed with the context items showed an assimilation effect, whereas those who disagreed showed a contrast effect. It is concluded that due to the inherent instability and complexity of attitudes, responses to attitude questions are likely to be influenced by the immediate context in which they are asked.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.418026  DOI: Not available
Share: