[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Stata for ERP data [Re: st:: question]
Kathy Kessler asked about the suitability of Stata for data analysis in the areas of
psycholinguistics and neuroscience, specifically electrophysiology.
_________begin excerpt from posting___________
We will be doing basic repeated measures MANOVA, with fairly complicated ERP data,
using planned comparisons. I don't believe this is anything out of the ordinary, though I
am fairly new to this type of analysis.
__________end excerpt from posting_____________
Stata in Release 8 has a built-in set of commands for MANOVA. They allow specifying
constrast matrixes just as in SAS and SPSS, so that Stata can handle, for example,
doubly multivariate MANOVA, treatment group-by-time interactions and the like.
In addition, the package has a wealth of time-series analysis methods. (The stand-
alone user's manual for its suite of time-series commands runs nearly 350 pages.)
These are primarily described in the terminology of econometricians (the target
audience), and I am not familiar with the majority of the techniques. But some of these,
for example, multivariate versions of time-series methods such as the vector
autoregression techniques that are new in Release 8, seem as if they might be more
powerful for analyzing experimental event-related/evoked-response potential (ERP)
data and other electrophysiological data in their native time-series format than
subjecting summarization parameters from them (P300 and so forth) to MANOVA.
* For searches and help try: