On Wed, 15 Jan 2003, Conn, Judith wrote: > I will be working with a longitudinal data set and do not know the > longitudinal methods. Could someone point me to an fairly easy to > read text or papers on these methods? Thank you, Judy Conn . .
Hi, Judy. A useful answer depends partly on what you have in mind by "longitudinal". If you have lots of time points (especially, but not necessarily, if they're equally spaced in time), you may want to consult texts on time series analysis. (Experts in such analyses -- which category doesn't include me! -- can tell you how long a series you'd need for this to be worth your while.) OTOH, if you're looking at data where the different times represent, say, baseline_measurement / pretest / posttest / follow-up_#1 / follow-up_#2 (or even this sort of thing repeated for two or three subsequent years), consult texts on repeated-measures ANOVA design and analysis, both univariate and multivariate forms of analysis. The current edition of Keppel (Design & analysis: A researcher's handbook; Prentice-Hall) may be useful, especially if you're dealing with psychological/educational/social-science data. Keppel calls designs of this type "within-subjects designs". Whether repeated-measures techniques are readily applicable depends also on whether you can track, from one time to the next time, the behavior of individual entities (persons?). If all you have is summary data, or even if you have individual scores but can't tell which are the same individuals from one time to the next, much of the sensitivity of repeated-measures analysis (which depends on a positive correlation of pretest with posttest (etc.)) will not be available to you. Good luck! -- DFB. ----------------------------------------------------------------------- Donald F. Burrill [EMAIL PROTECTED] 56 Sebbins Pond Drive, Bedford, NH 03110 (603) 626-0816 . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
