Commands to estimate models with variables at different levels?

[Cooper Felix]

Hi folks, imagine I have a panel data set in which the dependent variable is continuous and is measured at daily level, one predictor is measured at daily level, one is measured at weekly level, while the last one is measured at quarterly level. In this case, besides using xtmixed, what other commands do you frequently use? Thanks.

[Phil Bromiley]

There is nothing that stops you from including all such variables in an analysis of your daily outcome variable. If this is a panel, then xtreg is likely reasonable. The trade off is partially whether you need more levels of fixed/random effects.

[Marcos Almeida]

If you’re using a new version of Stata, instead of xtmixed you’re supposed to use - mixed - command.