Announcement

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

  • Propensity score matching

    Hi,
    Our outcome variable is birthweight and independent variables are antenatal care, education, age
    We want to apply PSM on it and we are looking for the commands on that since we are having a trouble running the commands

    Here is our sample data set

    Code:
    * Example generated by -dataex-. To install: ssc install dataex
    clear
    input double birth_weight byte(antenatal_care age edu)
        2 2 22 2
        4 1 35 5
        3 1 34 4
    2.724 2 35 2
        3 3 30 2
        2 2 33 7
     2.25 6 34 2
      2.5 1 37 3
        4 1 32 2
    9.998 3 26 3
        5 1 23 6
      2.3 1 25 3
        1 4 24 .
      3.5 1 24 .
    9.998 3 34 .
        3 5 22 .
     .898 2 32 2
      2.5 1 20 2
    9.998 4 29 4
        3 2 20 2
    9.998 5 37 .
        2 4 31 3
    9.998 7 30 2
        1 2 26 .
        2 3 33 3
        5 6 29 .
        3 6 25 2
    9.998 6 31 .
    1.635 1 37 6
      2.5 6 37 .
      .98 1 28 6
    1.362 4 35 .
        3 1 36 .
     2.05 4 46 .
    9.998 6 40 5
     2.11 7 35 .
        6 2 22 1
        3 3 35 3
        3 1 34 2
        1 2 24 3
        3 3 28 5
        3 4 40 4
        1 5 30 .
        3 1 40 .
      3.5 1 30 .
      3.2 2 27 .
      1.5 1 30 .
        3 2 28 2
        2 3 25 .
        3 7 27 2
      2.5 2 19 3
    9.999 1 22 2
        2 3 20 4
        1 7 27 3
      1.5 6 25 .
      1.5 3 31 2
        3 3 28 2
      1.5 4 34 1
    9.998 6 28 2
        3 1 25 .
        3 4 20 .
      1.5 4 28 3
        3 3 27 2
        3 2 25 5
    9.998 3 20 .
        5 1 47 2
    9.999 2 28 4
     .908 2 29 4
        3 2 27 6
        2 1 24 .
    9.998 3 34 .
      2.5 6 34 3
    9.999 6 33 .
      2.5 4 34 4
      2.5 2 40 5
    1.362 6 25 .
        3 1 40 3
        2 1 24 2
    1.136 2 27 2
    9.998 3 22 .
        2 2 28 .
        3 4 34 5
        4 2 37 .
     2.27 3 29 .
    1.362 2 23 3
    9.998 3 28 2
        4 1 28 2
      3.3 4 30 5
        4 3 27 6
     .908 3 32 6
     2.27 7 32 2
        1 1 19 .
    9.998 3 23 2
    1.454 3 18 .
        3 1 18 3
        2 1 29 2
        5 6 38 .
     .908 2 30 3
     .908 4 29 4
        3 2 27 4
    end
    label values birth_weight labels136
    label values antenatal_care labels85
    label values age labels17
    label values edu labels20
    Looking forward to your guidance in this regard.

  • #2
    It appears your treatment variable is antenatal_care. It appears to take on at least the values 1 through 7, so it is multiple treatments. How is the treatment measured? What would you like the base comparison group to be?

    PS matching is not as clear cut when you have more than two treatment levels, but it can be done pairwise.

    Comment


    • #3
      Thank you Jeff Wooldridge
      kindly guide me about PS matching in case of only two treatment levels (https://www.statalist.org/forums/for...score-matching)
      Last edited by Moomal Khan; 21 Aug 2022, 16:26.

      Comment


      • #4
        See https://www.statalist.org/forums/for...93#post1513193 and
        https://www.statalist.org/forums/for...08#post1242208
        David Radwin
        Senior Researcher, California Competes
        californiacompetes.org
        Pronouns: He/Him

        Comment

        Working...
        X