The Reliability of Survey Attitude Measurement
ABSTRACT Several theoretical hypotheses are developed concerning the relation of question and respondent characteristics to the reliability of survey attitude measurement. To test these hypotheses, reliability is estimated for 96 survey attitude measures using data from five, 3-wave national reinterview surveys-three Michigan Election Panel Surveys and two reinterview studies conducted by the General Social Survey. As hypothesized, a number of attributes of questions are linked to estimated reliability. Attitude questions with more response options tended to have higher reliabilities, although there are some important exceptions. More extensive verbal labeling of numbered response options was found to be associated with higher reliability, but questions explicitly offering a “don't know” alternative were not found to be more reliable. Question characteristics were confounded to an unknown degree with topic differences of questions, which were significantly linked to reliability, leaving the influence of question characteristics on reliability somewhat ambiguous. Characteristics of respondents were also found to be related to levels of reliability. Older respondents and those with less schooling provided the least reliable attitude reports. These results are discussed within a general framework for the consideration of survey errors and their sources. Peer Reviewed http://deepblue.lib.umich.edu/bitstream/2027.42/68969/2/10.1177_0049124191020001005.pdf
Full-textDOI: · Available from: Duane F. Alwin, Mar 18, 2015
- SourceAvailable from: Mark Shevlin[Show abstract] [Hide abstract]
ABSTRACT: A series of Monte Carlo simulations were carried out to examine the performance of Cronbach’s alpha as an index of reliability. Data were generated to be consistent with a single factor measured with six items. The magnitude of the factor loadings, systematic error and sample size were manipulated and alpha calculated from random samples. The results showed that alpha is influenced by factors other than the reliability of the items that comprise a scale. In particular the amount of systematic error, or deviation from unidimensionality, increased the estimate of alpha. The results are discussed in terms of traditional interpretation of alpha.Personality and Individual Differences 02/2000; 28(2-28):229-237. DOI:10.1016/S0191-8869(99)00093-8
- [Show abstract] [Hide abstract]
ABSTRACT: Existing methodologies of bank branch performance analysis are dominated by accounting and financial measures that inherently generate information of a retrospective nature. The outcomes are a possible misleading assessment of a branch's economic viability and misguided planning decisions in reconfiguring a branch. ... This study develops three principal models for an integrated analysis of personal banking performance, based on data collected from branches of a major Australian trading bank. In conclusion, the study delivers a model of overall performance comprised of eight variables; a model of overall potential also comprised of eight variables,- and, a discriminant model comprised of five predictors and two functions with validated coefficients, and a cross-validated correct classification rate. In addition, scales developed for such constructs as customer service quality and managerial competence can be applied independent of regression analysis and discriminant analysis models. It is submitted that the findings of the study can be used in decisions concerning reconfiguring, closing, or opening branches. Furthermore, the study could assist in minimising the gap between current branch performance and branch potential, by focusing attention on the treatment of variables that are controllable by bank management.