Display mean and median test results in Stata

Sometimes we may want to produce the following table to compare the mean and median of two groups:

First of all, please refer to this post to see Stata commands to test equality of mean and median.

However, it is time-consuming to glean numbers from the output of these Stata commands and place them in a table. It is even more struggling that you have to repeat the tedious process every time you update your sample.

I write Stata codes to streamline the process. The codes vary between unpaired (i.e., unmatched) data and paired data.

Unpaired data

The above example is unpaired data, i.e., suspect firm-years and other firm-years are not 1-to-1 or 1-to-m matched. One usage of unpaired data is the first step of Heckman’s two-step procedure, in which two groups of observations (i.e., the group that will be selected into the second step and the group that will not be selected into the second step) are stacked vertically in the dataset. The following codes are used for unpaired data. You only need to modify the first two lines to suit your data. The codes will generate a table in Stata’s output window like this:

You can then select the output and right-click “Copy as table” and paste in Excel for a quick edit. The codes use t-test for mean and Wilcoxon rank-sum test for median.

Paired data

A typical usage of paired data is to identify a matched control group for the treatment group. For example, identify a matched firm-year for an event firm-year based on a set of characteristics (same industry, same year, similar size and book-to-market), or identify a matched firm for every event firm based on the closest propensity score (i.e., propensity score matching).

The following table is an example that compares the mean and median of two matched groups—restating firms and non-restating group. Each restating firm is matched with a non-restating firm.

Because of this matching relationship, every event firm and its control firm will be placed in the same row in the dataset. In other words, event firms and control firms are aligned horizontally. The following codes are used for paired data. You only need to modify the first two lines to suit your data. You must specify the same number of variables in the matched order in the first two lines. In other words, the first variable in the first line must be paired with the first variable in the second line, and so on. The codes will generate a table in Stata’s output window like this:

The codes use paired t-test for mean and Wilcoxon rank-sign test for median.

 

This entry was posted in Stata. Bookmark the permalink.

22 Responses to Display mean and median test results in Stata

  1. HP says:

    Thank you for much for sharing this program.

  2. Victor says:

    You can also directly output the results to excel by using putexcel.
    Just change the display to something like, putexcel A1 = “Mean” , where A1 is cell in Excel.

    Some instructions:
    https://blog.stata.com/2017/01/24/creating-excel-tables-with-putexcel-part-2-macro-picture-matrix-and-formula-expressions/

    • Kai Chen says:

      Thanks for your comments. However, this type of test involves extracting returned scalars from MULTIPLE commands and presenting them in a single table. putexcel works great if you only need returned scalars from a SINGLE command. But I am afraid it cannot pick up results from multiple commands and export them to an Excel.

  3. Pedro Coelho says:

    Thank you very much for the command.

    Is there a way to include significance stars in the diferences (like the example from Zang, at the beginning of the post)?

  4. George says:

    Hi Kai: sorry I tested your code in Stata 14 but it was producing empty results. Is your code working for only 4 variables.

    • Kai Chen says:

      Hi George, it works for >4 variables. But you have to organize your data in certain ways and let the code know the required group variable. For example, unpaired data should be stacked by group and paired data should not.

  5. Vincent says:

    Hi Kai:
    is it possible to modify your code to include quantile regression for medians. According to Stata Journal qreg is best to use for differences in medians. https://journals.sagepub.com/doi/pdf/10.1177/1536867X1201200202

  6. George says:

    Thanks Kai: how can we export your results into Excel format.

  7. Kristin says:

    Thanks Kai! Your code helped me get a nice table in the State output window. I’m wondering whether you have any tips for getting the table from there to Word or Excel?

    • Kai Chen says:

      Please select the output table in Stata output window. Right-click and select “Copy as Table”. Then paste it in an Excel sheet and do format there and then move it to Word.

  8. selma boussetta says:

    Hi Kai,

    First of all thank you very much for sharing your code and for your helpful explanations.
    I would apply your code but i have two dummies one for treated/control group and one for before/after the event. Could you please help me modify this code ? My data is not matched by pairs. Thank you again

  9. Vas0 says:

    Hello everyone!!
    Thank you very much for sharing.

    I am trying to run the codes for unpaired data. My group variable is MNC (multinational) dummy.

    However, when I run the codes I get the following error:
    noi di _col(15) “{c |} `group’ = 1” ///
    | MNC = 1invalid syntax
    .
    It is the only command that does not run so I cannot get the final table.

    Do you have any idea on how I can overcome this issue?

    Thank you very much for your time and consideration.

    • Kai Chen says:

      Just change the first two lines and leave other codes unchanged. The first line will be the variables that you want to compare. Change the second line to: local group MNC.

      • Vaso says:

        I have already changed the first two lines as you indicated in the beginning, but still there is nothing that comes to my mind that I might be missing.

        Thank you 🙂

        • Kai Chen says:

          I mean, change the first two lines only. You also changed this line: noi di _col(15) “{c |} `group’ = 1” ///
          | MNC = 1. Do not add | MNC = 1 at the end.

          • Vaso says:

            When I run this:
            noi di _col(15) “{c |} `group’ = 1” ///

            I get this error—->| MNC = 1invalid syntax | MNC = 1<—–

            I will update you if I figure it out. Thank you for your quick responses.

  10. R Zhang says:

    Hi Kai,

    Thanks heaps for sharing the code! You are a life saver!

    I experienced the same problem as Vaso. I had 3 investor types – ethical, interested and conventional. I dropped all observations in the conventional group and was hoping to compare ethical investors (GROUP = 1) and interested investors (GROUP = 0). I changed the first two lines only, everything worked perfectly fine until I got an error message “GROUP = 1invalid syntax”. Not sure why this happened… I was wondering if I have missed something…

    Thanks again for your help!

  11. Dee says:

    Hi Kai,

    Thanks for your help and sharing the code freely!

    I have a question related to the result, how can i get the *(star marks) that indicate signifficance similar to the example that you provide in the beginning of post?

    Thanks in advance!

  12. Oliver says:

    Dear Kai,

    Thank you so much for sharing this code. Unfortunately, I have the same problem as R Zhang and Vaso. I have only two groups (high and low), but I also receive the massage:

    noi di _col(15) “{c |} `group’ = 1” ///
    | HighAcquirerESGtotalscore = 1invalid syntax

    Did you find the problem in the data from R Zhang and do you maybe have a general answer to solve this problem?

    Thanks again for your help!

  13. Narayanan Venkatraman says:

    Hi Kai,
    noi di “{hline 115}”
    ——————————————————————————————————————-
    . noi di _col(15) “{c |} `mygroup’ = 1” ///
    | TREAT //invalid syntax

    I am having the same issue with the code. As suggested I modified only the first two lines of the code. My group variable is TREAT.

Leave a Reply to selma boussetta Cancel reply

Your email address will not be published. Required fields are marked *