Which method can be used to analyze daily panel data in combination with annual survey data (on income, attitudes towards organic products and so on)? The method should also be able to measure changes over time (e.g., in attitudes toward organic products and actual purchases of organic products).
Possible research question: How do consumer attitudes (annual measrured by asking the consumers once a year) and actual purchasing behavior (daily measured) change over a long period of time?
I have several papers on time-dependent ARMA and VARMA models (and more on the way) but it seems difficult to me to relate daily to annual data. Good luck.
Hi Guy Mélard
Thanks for your answer! I will check your mentioned publications!
Could the combination of annual and daily data perhaps be possible if I first aggregate the purchase data into calendar weeks and then use the individual survey data of the people who bought in that calendar week? So I could calculate per calendar week, for example, the sales and attitudes of the people who bought in that calendar week.
But are there approaches to analyse this panel data aggregated to calendar weeks with the help of time series analyses? Or should time series analyses tend to be done at the individual level (days or households)?
Thank you very much!
I have an unpublished paper on daily and high-frequency data. I just added it to ReserchGate with
DOI: 10.13140/RG.2.2.36742.70728
You will see that I had similar problems but I didn't cover panels on which I have no practice.
Dear Phillipp,
I will recommend ARMA and VERMA @Guy Melard since he has worked in that area.
Then you will need a focasting model to answer the questions for the future.
Regards
- Badges
- Science topic
- Similar topics
- Chemistry, Organic
- Organic