# Re: st: ttest and log transformation

 From Nick Cox <[email protected]> To [email protected] Subject Re: st: ttest and log transformation Date Mon, 29 Sep 2008 09:40:23 -0500

I want to back up and ask what you want to do, and what you think the t test would do for you.

It is a big jump from a very general question like

How do these distributions differ?

to a specific question like

Are the means of these distributions the same, or different?

or

How do the means of these distributions differ?

The second and third require at a minimum that means are useful for your data. t tests are not general answers to the first.

A useful direct way to compare two distributions is through -qqplot-. That can give you a direct signal on whether two distributions differ by an additive shift, which (with other stuff) lies behind the t-test, or by a multiplicative shift, which (with other stuff) lies behind a t-test on logged values, or, as I suspect, something much more complicated.

What best to do with data that can be + or -, are long-tailed and skew in one direction is not, it seems, often discussed, although it is exactly what I would expect with say company profit and loss data, which are hardly exotic. The help file -transint- on SSC has some discussion on a neglog transformation.

Nick
[email protected]

Richard Harvey wrote:

```I hope I can ask a fairly basic stats question. I have a variable that
i need to compare across two groups.
the summary stats for the variable NAN  across the groups is as below.
The negative values are legitimate.

group   |            N             mean             p50           max
min                skewness  kurtosis

group1 |           2537         -77535           5278       19051350
-46844688         -11.23          311.1
group2 |           3031        -211373           4620        4609996
-32617714         -11.18          185.6
Total   |          5568        -150391           4958       19051350
-46844688         -11.33          278.4

If a do a ttest on the log transformed data, is it appropriate to add
an arbitrary constraint to make the negative values positive?  Is the
ttest indeed any good for this data, or should I be looking at some
non parametric tests.

to make the numbers more manageble is divide by 1000,000 and the
summary stats look like this

group	          N		mean	p50	                max	           min	skewness	kurtosis

group1		2537		-.07753	.005278		19.05	-46.84	-11.23	311.1
group2		3031		-.2114	.00462		4.61	        -32.62	-11.18	185.6
Total		        5568		-.1504	.004958		19.05	-46.84	-11.33	278.4

Is it right to perform ttest on ln((NAN/1000000)+50) ? changing the
constant i add dosent seem to make a difference.

stats on ln((NAN/100000)+50) is as below

group	             N		mean	p50	                max	         min
skewness	kurtosis

group1		2537		4.604	4.605		4.78	       3.973	            -17.21	527.4
group2		3031		4.603	4.605		4.65	         4.21	             12.74	242.9
Total		        5568 	4.604	4.605		4.78	       3.973             -15.94	469

There is still a large negative skewness coefficient.  To me this
looks like not a situation for a  ttest and I should be looking at
some non parametric test. Is that right?

The results from the ttest using the unpaired and unequal option,
using the untransformed and using ln((NAN/100000)+50) are as below

transformation               t                 p                       95% CI
None                          3.25            .0011
53205.45-214470.8
log(50+var)                 2.75            .0060
.000367 - .002185 ( I understand this has to be back transformed)

a ranksum test on the logtransformed NAN shows a z of 3.3999 with a p
of .0007.on the untransformed NAN it is 3.396 with p of .0007

so overall, there dosent seem to be any change in the conclusions,
what ever test I use. But is the ttest procedure appropriate?

You help is much appreciated.
```
```*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```