Lower 5th quantile of a variable

[Aamina Khurraa]

Hi,
I am new to the forum. I am working on conducting some statistical analysis in stata. I have couple of queries:

1. I am very confused in differentiating between centile, xtile and p in stata. I want to find the lower 5th percentile (quantile) of a variable for each panel ID that I have (by ID). please help me know the relevant code.

2. I am conducting quantile regression on a panel data set and I want to know the difference between "xtqreg" and "by panelID qreg". Also when I want to run the regression individually for each Panel ID that I have tried to do using xtreg, it doesnot accepts "by" option and gives the aggregate quantile regression results for the entire data-set and not just the individual firms or panel IDs.

3. Lastly, when I use "by panelID qreg" and want to output the results in asdoc, it doesnot considers sorting w.r..t "by panelID" rather only gives the aggregated regression result for the entire data-set. How can I output regression results using asdoc for individual firms using "by panelID" or "bysort PanelID"

Much Thanks
Aamina

[Nick Cox]

I will peel off Q1. I note that in Q2 and Q3 you refer to communitty-contributed commands xtqreg and asdoc. You're asked to explain where they come from (https://www.statalist.org/forums/help#stata).

Q1. You can put the 5th percentile (sometimes described as the 0.05 quantile) in a variable with something like

. webuse grunfeld , clear

. egen p5 = pctile(invest), p(5) by(company)

. tabdisp company, c(p5)

----------------------
  company |         p5
----------+-----------
        1 |     287.65
        2 |     220.15
        3 |      38.85
        4 |     43.545
        5 |     39.675
        6 |      22.48
        7 |      23.82
        8 |     15.885
        9 |      22.14
       10 |      1.055
----------------------

. egen p5y = pctile(invest), p(5) by(year)

. tabdisp year, c(p5y)

----------------------
     year |        p5y
----------+-----------
     1935 |       2.54
     1936 |          2
     1937 |       2.19
     1938 |       1.99
     1939 |       2.03
     1940 |       1.81
     1941 |       2.14
     1942 |       1.86
     1943 |        .93
     1944 |       1.18
     1945 |       1.36
     1946 |       2.24
     1947 |       3.81
     1948 |       5.66
     1949 |       4.21
     1950 |       3.42
     1951 |       4.67
     1952 |          6
     1953 |       6.53
     1954 |       5.12
----------------------

I don't think I can improve on the help and manual entries in explaining centile and xtile. Sorry, but I have no idea what you mean by p. You will note that neither of the first two commands is what I recommend here.

[Aamina Khurraa]

Thank you Nick for your valued assistance.

Regarding Q2 and Q3, if I am not getting your question wrong, I am using xtqreg in stata 15.1. which is quantile regression for panel data. The problem is that I can't find any code to perform the quantile regressions at 0.05 for individual firms or panels in the data.

Best
Aamina

[Nick Cox]

The point is that xtqreg and asdoc are not official commands, but on SSC. If you don't get replies -- authors of each are members here but may not read this thread -- you may need to start new threads mentioning those commands.

[Bruce Weaver]

Alternatively, you try something like this to attract the attention of the authors of those packages to this thread:

xtqreg - written by J.A.F. Machado & Joao Santos Silva
asdoc - written by Attaullah Shah

Members you name using the @name feature will be notified that they have been mentioned in a post.

[Aamina Khurraa]

Thank you Nick and Bruce. I will follow

[Attaullah Shah]

Thanks, Bruce Weaver for tagging me in this post. Since xtqreg is a community-contributed program, asdoc was not aware of that. I have added this to the regression group, and now asdoc can treat it like any other Stata regression command. The new version of asdoc can be installed from my site. Copy and paste the following line in Stata and press enter.

net install asdoc, from(http://fintechprofessor.com) replace
sysuse auto, clear
asdoc xtqreg price weight length i.foreign, i(headroom) quantile(.25)

Please note that the above line has to be copied in full. After installation of the new version, then restart Stata.

Please do remember to cite asdoc. To cite:
In-text citation
Tables were created using asdoc, a Stata program written by Shah (2018).

Bibliography
Shah, A. (2018). ASDOC: Stata module to create high-quality tables in MS Word from Stata output. Statistical Software Components S458466, Boston College Department of Economics.

[Aamina Khurraa]

Thank you so much Attah Ullah Shah. I look forward to further assistance from you all.

I will definitely try this and cite asdoc in the paper

Best
Aamina