Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • no observation error (2000) after entering control variables.

    Hi, I'm working on mexico pension program and at primary steps, when I want to estimate impact of program by linear regression it works but when I enter my control variables I get below results in the picture, (no observation). I read in Other topics that it can be because of string variables and omitted those but still get same error. for information I have missing data if it is a case, I read in other topic that maybe missing data can be an issue but I don't know how to resolve it. thanks.

    Click image for larger version

Name:	mexico.png
Views:	1
Size:	2.1 KB
ID:	1547894


  • #2
    Code:
    su spending_doc time $controls id_loc
    will tell you how many non-missing values you have for each, but you need observations in which all variables are non-missing -- and it seems you have none.

    Comment


    • #3
      Thanks, I ran the command and it had a good information on the results and as you said it shows non of my control variables have full set of observation. Can I put 0 instead of missed data? does it cause an intervention error in original data set?

      Comment


      • #4
        i even tried use looping command as it was suggested in other topic but it gives same error.
        . foreach x of global controls {
        2.
        . describe `x'
        3.
        . reg `x' spending_dec treatment $controls, cl(id_loc)
        4.
        . }

        Comment


        • #5
          Missing means zero only if missing means zero!

          Comment


          • #6
            Thanks Nick, then this option did not approved as well and I should ignore it.

            Comment

            Working...
            X