An unfolding latent variable model for likert attitude data: Drawing inferences adjusted for response style

• 成果类型: Article
• 署名作者: Javaras, Kristin N.; Ripley, Brian D.
• 署名单位: University of Oxford
• 刊物名称: JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
• ISSN/ISSBN: 0162-1459
• DOI: 10.1198/016214506000000960
• 发表日期: 2007
• 页码: 454-463
• 关键词: cross-cultural research; extreme; sets

摘要

Likert attitude data consist of responses to favorable and unfavorable statements about an entity. where responses fall into ordered categories ranging from disagreement to agreement. Social science and rnarketing researchers frequently use data of this type to measure attitudes toward an entity such as a policy or product. We focus oil data on American and British attitudes toward their respective nations (national pride). We introduce a multidimensional Unfolding model (MUM) to describe the relationship between the data and the attitudes underlying I Just attitudes. but also response style, which is defined as a them. Unlike most existing, models, the MUM allows the data to reflect not consistent and content-independent pattern of response category selection such as a tendency to agree with all statements. The MUM can be used to model multiple attitudes, which allows researchers to expand their analysis of the data of interest to include all available Likert data so as to increase information on response style. For example. we include additional data on immigration attitudes to help distinguish the effects of response style and national pride on our data. The MUM can be used to fit linear models for the effects of background variables oil attitudes. Resulting inferences about attitudes are adjusted for response style arid should be less biased. Simulation results strongly suggest that, unlike Likert's popular scoring model. the MUM yields unbiased inferences even when there are Unequal proportions of favorable and unfavorable statements.