Support #1462: respondents without a record in the jk data weights and representativeness - Understanding Society User Support

Actions

Copy link

Support #1462

open

respondents without a record in the jk data weights and representativeness

Added by Laura Jones about 4 years ago. Updated almost 4 years ago.

Status:

Resolved

Priority:

Normal

Assignee:

Category:

Weights

Start date:

12/10/2020

% Done:

100%

Description

Dear US team,
I don't believe this question has been asked elsewhere or addressed in the user manual, but please do let me know if I'm wrong.

I want to analyse date from the July wave and use information from the jk dataset to construct certain variables measured pre-covid that are unavailable in the covid surveys. There are 589 respondents to the July covid survey without records in the jk dataset.

I know that the cross-sectional weights account for non-response between the survey and wave 9. Would I be correct in saying that they do not correct for differential response between wave 9 and the jk dataset? And that therefore by including variables measured in the jk dataset the weights would no longer provide estimates representative of the population 16+ at the time of wave 9. If so could you provide guidance on how variables in the jk dataset can be used to provide population estimates.

many thanks!
Laura

Actions

Copy link

Updated by Alita Nandi about 4 years ago

Status changed from New to In Progress
Assignee deleted (~~Olena Kaminska~~)
% Done changed from 0 to 10

Many thanks for your enquiry. The Understanding Society team is looking into it and we will get back to you as soon as we can.

We aim to respond to simple queries within 48 hours and more complex issues within 7 working days. While we will aim to keep to this response times due to the current coronavirus (COVID-19) related situation it may take us longer to respond.

Best wishes,
Alita
Understanding Society User Support Team

Actions

Copy link

Updated by Alita Nandi about 4 years ago

Status changed from In Progress to Feedback
Assignee set to Laura Jones
% Done changed from 10 to 80
Private changed from Yes to No

Hello Laura,

Here are our responses:

"Would I be correct in saying that they do not correct for differential response between wave 9 and the jk dataset?"
- Yes

"And that therefore by including variables measured in the jk dataset the weights would no longer provide estimates representative of the population 16+ at the time of wave 9."
- Yes

If so could you provide guidance on how variables in the jk dataset can be used to provide population estimates.
- The xd_betaindin_xw weight is still the best, albeit sub-optimal, option. These weights will not correct for non-random attrition between W9 & W10/11. But also note that of these 589 cases 320 have zero xd_betaindin_xw weights and so would anyway be excluded from your analysis. So, you would be actually be excluding only 269 observations from your analysis due to no jk_ data.

Hope this helps.

Actions

Copy link