Premium
Separating interviewer and area effects by using a cross‐classified multilevel logistic model: simulation findings and implications for survey designs
Author(s) -
Vassallo Rebecca,
Durrant Gabriele,
Smith Peter
Publication year - 2017
Publication title -
journal of the royal statistical society: series a (statistics in society)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.103
H-Index - 84
eISSN - 1467-985X
pISSN - 0964-1998
DOI - 10.1111/rssa.12206
Subject(s) - statistics , interview , estimator , multilevel model , variance (accounting) , sample size determination , econometrics , wald test , dispersion (optics) , sample (material) , logistic regression , set (abstract data type) , hierarchical database model , statistical hypothesis testing , mathematics , psychology , computer science , data mining , economics , physics , chemistry , accounting , optics , chromatography , political science , law , programming language
Summary Cross‐classified multilevel models deal with data pertaining to two different non‐hierarchical classifications. It is unclear how much interpenetration is needed for a cross‐classified multilevel model to work well and to estimate the two higher‐level effects reliably. The paper investigates this question and the properties of cross‐classified multilevel logistic models under various survey conditions. The effects of different membership allocation schemes, total sample sizes, group sizes, number of groups, overall rates of response and the variance partitioning coefficient on the properties of the estimators and the power of the Wald test are considered. The work is motivated by an application to separate area and interviewer effects on survey non‐response which are often confounded. The results indicate that limited interviewer dispersion (around three areas per interviewer) provides sufficient interpenetration for good estimator properties. Further dispersion yields only very small or negligible gains in the properties. Interviewer dispersion also acts as a moderating factor on the effect of the other simulation factors (sample size, the ratio of interviewers to areas, the overall probability and the variance values) on the properties of the estimators and test statistics. The results also indicate that a higher number of interviewers for a set number of areas and a set total sample size improves these properties.