ESTIMATING INTERINTERVIEWER RELIABILITY FOR INTERVIEW SCHEDULES USED IN SPECIAL-EDUCATION RESEARCH

Using data from an individually administered interview schedule, the Consumer Satisfaction Inventory (CSI), interinterviewer reliability was estimated with several different approaches: simple percentages of agreement, kappa and weighted kappa, Pearson correlations, t tests on interviewers’ means, and generalizability (G) theory techniques. The reliability estimates varied, sometimes widely. Differences and similarities among the approaches are discussed, and some suggestions are given to aid researchers in choosing estimation techniques for particular situation. © 1991, SAGE Publications. All rights reserved.