Inconsistencies in self-completion module of wave three?
In wave three of the Understanding Society survey, there are six entries for respondents who have refused the self-completion part of the interview but have positive self-completion interview weights. In particular, c_csac is 4 (refused) or 5 (not able to complete) (see https://www.understandingsociety.ac.uk/documentation/mainstage/dataset-documentation/wave/3/datafile/c_indresp/variable/c_scac) but c_indscub_xw is nonzero (see https://www.understandingsociety.ac.uk/documentation/mainstage/dataset-documentation/wave/3/datafile/c_indresp/variable/c_indscub_xw).
Similarly, there are three entries indicating that the self-completion part of the interview was refused but the response to the "What is the sex of your first friend?" question contains valid data even though the question is in the self-completion module.
Am I missing something?
#1 Updated by Olena Kaminska over 4 years ago
Thank you for letting us know. You are right that the 6 cases mentioned should have 0 cross-sectional self-completion weight. The issue occurred most likely later at the data release stage (and after the weights were created) when updated information suggested that some self-completion interviews were not valid. The number of non-zero weights is very small and this should not influence your analysis - just use the provided weights in the normal way.
Thank you again for pointing the issue to us,
#2 Updated by Alita Nandi over 4 years ago
- Assignee set to Alita Nandi
- % Done changed from 0 to 40
You are correct. There are a handful of cases for whom SCAC shos that they "refused self-completion" or "not able to do self-completion" but yet have some valid self-completion data. We are looking into this. For now, you should exclude these cases from your analysis.
#3 Updated by Till Hoffmann over 4 years ago
Dear Olena and Alita,
Thank you for the updates. I will omit these cases for now.
It also appears that there are 2477 responses who have agreed to complete the self completion module (c_scac >= 1 and c_scac <= 3) but have zero cross-sectional weight c_indscub_xw.
#5 Updated by Olena Kaminska over 4 years ago
Thank you. With regard to people who completed self-completion questionnaire but have 0 weights: this issue reflects that UKHLS is a longitudinal and not a cross-sectional study. The weights reflect how people were selected and how they responded over the years. Most of the cross-sectional weights are based on non-zero psnenub_xw weight. Zeroes in self-completion weight reflect zeroes in the enumeration weight.