I am using a dataset that has three waves divided by months (e.g., Wave 1 April to June 2021, Wave 2 July to September 2021, Wave 3 October to December 2021). Some of the participants have participated in just one wave while others are in all three.

I defined the analyses I am doing as repeated cross-sectional analyses as I am examining each wave separately, but I have pooled the results and discussed the trends over time if that makes sense. However, I was also wondering if "cross-sectional time-series analysis" (or something equivalent) is be more fitting.

I am aware that there are variations in terminologies. For example, what econometricians and social scientists call "panel data", statisticians and epidemiologists refer to as "longitudinal data". Thus, the above distinction might be irrelevant.

  • 1
    $\begingroup$ Is the sampling scheme conducted de novo with each new wave? $\endgroup$ Jun 1, 2022 at 5:20
  • 1
    $\begingroup$ @ThomasBilach The second and third waves had participants from who also participated in the first wave. However, the cross-sectional weights at each wave are independent. $\endgroup$
    – guaguncher
    Jun 3, 2022 at 19:21


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy