Sociology, Department of
Date of this Version
2020
Document Type
Article
Citation
Appendix to Chapter 16, from Olson, K., J. D. Smyth, J. Dykema, A. L. Holbrook, F. Kreuter, B. T. West. 2020. Interviewer Effects from a Total Survey Error Perspective. Boca Raton: CRC Press.
Abstract
Appendix 16A: Descriptive statistics for available call records
Figure A16A.1 Histograms of wave 2 nurse visit call length by call status with outlier of 691 minutes removed from the ‘any interviewing done’ category
Figure A16A.2 Histograms of wave 3 nurse visit call length by call status with outlier of 464 minutes removed from the ‘any interviewing done’ category
Figure A16A.3 Bar chart of total number of calls per household at UKHLS wave 2 nurse visit
Figure A16A.4 Bar chart of total number of calls per household at UKHLS wave 3 nurse visit
Figure A16A.5 Histogram of total nurse visit time in minutes – wave 2
Figure A16A.6 Histogram of total nurse visit time in minutes minus the blood sample modules – wave 2
Appendix 16B: Full Results from Regression Models Corresponding to Tables 16.4-16.12 in Main Body
Table A16B.1 Results from cross-classified multilevel logistic regression models for the log-odds of participating in the wave 2 nurse visit using nurse characteristics and paradata variables. Corresponds to table 16.4 in main body
Table A16B.2 Results from cross-classified logistic regression models of the log-odds of consenting to the blood sample in the wave 2 nurse visit using nurse characteristics and paradata variables. Corresponds to table 16.5 in main body
Table A16B.3 Results from cross-classified logistic regression models of the log-odds of obtaining the blood sample in the wave 2 nurse visit using nurse characteristics and paradata variables. Corresponds to table 16.6 in main body.
Table A16.4 Results from cross-classified logistic regression models of the log-odds of participating in the wave 3 nurse visit using nurse characteristics and paradata variables. Corresponds to table 16.7 in main body.
Table A16B.5 Results from cross-classified logistic regression models of the log-odds of consenting to the blood sample in the wave 3 nurse visit using nurse characteristics and paradata variables. Corresponds to table 16.8 in main body.
Table A16B.6 Results from cross-classified multilevel logistic regression models of the log-odds of obtaining the blood sample in the wave 3 nurse visit using nurse characteristics and paradata variables. Corresponds to table 16.9 in main body.
Table A16B.7 Results from cross-classified logistic regression models of the log-odds of participating in the wave 3 nurse visit using nurse performance indicators. Corresponds to table 16.10 in main body.
Table A16B.8 Results from cross-classified logistic regression models of the log-odds of consenting to the blood sample in the wave 3 nurse visit using nurse performance indicators. Corresponds to table 16.11 in main body.
Table A16B.9 Results from cross-classified logistic regression models of the log-odds of obtaining the blood sample in the wave 3 nurse visit using nurse performance indicators. Corresponds to table 16.12 in main body.
Appendix 16C: Quantile-Quantile plot of nurse residuals for the model of having call records in wave 2
Figure A16C.1 Quantile-Quantile plot of nurse residuals with 95% confidence intervals from the model of having call records in wave 2. The green oval indicates the nurses with above-average call recording and the red oval indicates the nurses with below-average call recording.