Hello together,

I am curious if you have any concerns regarding my Descriptive Statistics?
In my work I use propensity score matching to get a treatment and control group for firms which adopted ICT and those ones which did not.

Code:
Summary statistics
N St.Dev min max t-value
Sales 8239 1.15e+11 0 9.85e+12 3.066
LogSales 8236 2.937 1.946 29.918 526.573
GPM 8236 1890000 -1.71e+08 100 -1.039
LogGPM 7715 .707 -3.727 4.605 499.368
Inno 8960 1.17 0 4 53.717
EXP 8960 19.362 0 100 32.666
LaPr 2930 3.89e+07 -1.22e+07 1.32e+09 2.914
LogLaPr 2877 3.034 -.85 21.005 159.151
CapaUtili 3975 22.667 0 100 194.115
ICT 8960 1.575 0 10 84.696
FirmSize 8876 733.866 0 64000 9.765
LogFirmSize 8864 1.346 0 11.067 207.275
PowerOutage 8820 .375 0 1 208.374
Firmage 8642 15.235 2 171 160.939
LogFirmAge 8642 .485 .693 5.142 603.1
Ownership 8795 29.071 0 100 36.351
Country 8960 7.531 10 36 304.077
I also use the variables in their logarithm form when available and it turns out that the results are more significant, for example for LogSales compared to Sales.
Is there any good justification for this approach except 'because my results are significant now'?

Kind regards
Dominik