Dear Zhu Yifei

The fundamental disadvantage of cross-sectional studies is that the temporal relationship between result and exposure cannot be ascertained because both are evaluated at the same time. Among a zoo, for example, reproduction is more typically impeded in animals with stereotypies.

There is no universally accepted or established time range for cross-sectional research. Normally, we are urged to finish data collecting, data analysis, and research completion / publication as quickly as feasible.

Data from a cross-sectional survey are collected at a particular moment in time. A survey is conducted to a new sample of respondents at successive time periods to generate repeated cross-sectional data. In the case of an annual survey, this means that respondents in one year will be different from those in the previous year.

Best Regards

Dr. Fatemeh Khozaei

Fatemeh Khozaei

Thank you for your answer, but there are two reasons that give me hope for adding this variable to my model.

First, it's more like panel data than pooled cross-sectional data, as i take county as my research unit. Second, the time variance of the problematic variable is related to regional difference. Because this variable is from websites of province's Bureau of Statistics, and they only provide data of certain year chosen by themselves.

Thus I think this time variance may be dealt with when i put province fixed effects in my model.

If the variable you want to add can be affected with time. You can't add.. it will be considered a major limitation of the study.

If its the same variables and not affected by the time then you can. If its affected then you cant.

If its the same variables and not affected by the time then you can. If its affected then you cant.

Thanks to all answers above! Finally i decided to only analysis regions that has this variable collected in the same year, and to leave the rest for robustness check.