Mean difference test

[Naina Grover]

hello there ,
I have an unbalanced panel data of banks comprising of 21 public,42 foreign and 21 private banks from the year 2005 to 2018..I have created dummies for each group as welI .I have three questions
How can i calculate mean profitability for each group considering i have one panel sheet
How can i calculate mean profitability for each group by year?
How can i test the difference in mean of these groups.
Iam using stata 14 version
Please help
Thanks in advance

[Rich Goldstein]

for exact code, please show example data using -dataex- (see the FAQ); my guess is that -egen- with various options and "by" will do what you want; see

help egen

[Phil Bromiley]

Also look at ttest.

[Naina Grover]

Rich Goldstein sir,
My data is as follows
input int(bankcode Year) byte(PUBLICDUMMY PRIVATEDUMMY FOREIGNDUMMY) double Returnonassets_x000D_
1 2005 0 0 1 2.87
1 2006 0 0 1 3.43
1 2007 0 0 1 3.42
1 2008 0 0 1 3.84
1 2009 0 0 1 4.75
1 2010 0 0 1 5.58
1 2011 0 0 1 4.53
1 2012 0 0 1 7.05
1 2013 0 0 1 5.03
1 2014 0 0 1 4.22
1 2015 0 0 1 5.13
1 2016 0 0 1 3.07
1 2017 0 0 1 3.48
1 2018 0 0 1 4.17
2 2005 0 0 1 -2.57
2 2006 0 0 1 .35
2 2007 0 0 1 .23
2 2008 0 0 1 3.96
2 2009 0 0 1 2.4
2 2010 0 0 1 1.13
2 2011 0 0 1 1.12
2 2012 0 0 1 1.87
2 2013 0 0 1 1.4
2 2014 0 0 1 .56
2 2015 0 0 1 .57
2 2016 0 0 1 .67
2 2017 0 0 1 .54
2 2018 0 0 1 .6
3 2005 1 0 0 1.33
3 2006 1 0 0 1.42
3 2007 1 0 0 1.26
3 2008 1 0 0 1.32
3 2009 1 0 0 .9
3 2010 1 0 0 1.16
3 2011 1 0 0 1.11
3 2012 1 0 0 1.02
3 2013 1 0 0 .64
3 2014 1 0 0 .57
3 2015 1 0 0 .29
3 2016 1 0 0 -.33
3 2017 1 0 0 -.13
3 2018 1 0 0 -1.96
4 2005 0 0 1 .55
4 2006 0 0 1 1.45
4 2007 0 0 1 1.28
5 2009 0 0 1 -7.92
5 2010 0 0 1 -4.24
5 2011 0 0 1 1.67
5 2012 0 0 1 .2
5 2013 0 0 1 -3.23
5 2014 0 0 1 -3.46
5 2015 0 0 1 -2.22
5 2016 0 0 1 -.64
5 2017 0 0 1 .96
5 2018 0 0 1 1.27
6 2005 1 0 0 1.59
6 2006 1 0 0 1.38
6 2007 1 0 0 1.31
6 2008 1 0 0 1.16
6 2009 1 0 0 1.09
6 2010 1 0 0 1.39
6 2011 1 0 0 1.36
6 2012 1 0 0 1.19
6 2013 1 0 0 .99
6 2014 1 0 0 .29
6 2015 1 0 0 .38
6 2016 1 0 0 .28
6 2017 1 0 0 .08
6 2018 1 0 0 -1.46
7 2012 0 0 1 .05
7 2013 0 0 1 .1
7 2014 0 0 1 .94
7 2015 0 0 1 1.11
7 2016 0 0 1 .85
7 2017 0 0 1 .67
7 2018 0 0 1 .89

[Naina Grover]

Phil Bromiley, Can we apply anova on panel data.I have created categorical variable for my types of banks

[Rich Goldstein]

your data sample (#4) has no entries for Private but here is one strategy; first, note that if you just want to compare the means you don't need to calculate them; just

gen byte group=0 if PU==1
replace group=1 if PR==1
replace group=2 if F==1
xtset bankcode Year
xtreg Re i.group

if you had the third group I could show the test statement to get the comparisons ignoring the reference group; see

help test

if you really think you need to calculate the means in your first two questions in #1:

bys group: even double meanprofit=mean(Re)
bye group Year: even double meanprofityr=mean(Re)

added: I don't think that I fully understand what you want so I'm not sure the above is fully responsive