Jim Seward
> I have a database of scores of subjects who were given the same
> memory test on an annual basis. The data look something like:
>
> SUBJECT YEAR 1 SCORE YEAR 2 SCORE YEAR 3 SCORE.....
>
> Tom 15 16 14
> Dick 12 13 10
> Harry 15 14 15
> .
> .
> .
>
>
> I would like to determine if the subjects scores tend to increase
> or decrease with time. I was thinking of using linear regression,
> with YEAR as the x variable and SCORE as the y variable.
>
> In order to do that, I think that I have to recode the data
> thusly:
>
> SUBJECT YEAR SCORE
> Tom 1 15
> Tom 2 16
> Tom 3 14
> Dick 1 12
> Dick 2 13
> Dick 3 10
> Harry 1 15
> Harry 2 14
> Harry 3 15
>
> Does anyone know of a way to do this in Stata?
Yes. This is a -reshape-.
First off, your variable names possibly are
subject year1score year2score year3score
you might want to -rename- them to a standard
stub + suffix form:
forval i = 1/3 {
rename year`i'score score`i'
}
then you want
reshape long score, i(subject) j(year)
See [R] reshape and for further tricks
and examples
http://www.stata.com/support/faqs/data/reshape3.html
Nick
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
