Support #2297
open
Conducting quarterly analysis
60%
Description
Afternoon,
I would like to produce some analysis looking at the use of food banks / food insecurity / material deprivation using the (to be published) W15 data and previous waves.
I would like to be able to compare quarterly rates for those measures (where available) from 2020 to 2025. The microdata contains the correct variables (date of interview and household id) to allow me to merge these files.
To correctly do this would I merge the datasets from those time periods (Wave 11 - Wave 15) and then recode the _quarter variable to clearly outline what calendar year and quarter households were interviewed in.
Then for analysis apply the hhdenub weight? I wanted to mainly check that there werent any sampling or weighting considerations that would prevent me from doing this analysis using these steps ? This seems to be the advice for creating the calendar year datasets.
Best,
Tom
Updated by Understanding Society User Support Team about 2 months ago
- Category changed from Data analysis to Weights
- Status changed from New to Feedback
- % Done changed from 0 to 60
- Private changed from Yes to No
Dear Thomas,
Thank you for you question – it’s worth having a look at the weighting FAQ (see https://www.understandingsociety.ac.uk/wp-content/uploads/working-papers/2024-01.pdf), especially Q6, Q13, Q14–Q16, as some of your steps may need a rethink. In particular, the document gives specific guidance on how calendar-time analysis should be constructed (e.g. using w_month rather than interview date, recommended weights for calendar-period comparisons, and adjustments when pooling waves).
Let us know if this helps.
Best wishes,
UKHLS User Support Team