Last week I asked how I might analyze a time-series type of before- after design, one subject at a time (with due regard for the dubious charm of such a quasi-experimental design). For example, one might collect a week of baseline on something, then a week of data under treatment. I'd like to thank everyone who contributed to that thread, even though I eventually found the answer for myself (no, this wasn't a plant). The answer is to use a method known, among other names, as an interrupted time-series analysis, which calculates autocorrelations. In case anyone's interested, I've appended a number of references below. But the jewel in the crown is a paper I found by Crosbie (1993). He developed a variant of this analysis he calls ITSACORR, and provides a standing offer of a programme to calculate it. 48 hours later, the programme was up and running on my computer. It's really neat. -Stephen References Crosbie, J. (1993). Interrupted time-series analysis with brief single-subject data. Journal of Consulting and Clinical Psychology, 61, 966-974 Jones, R. et al (1977). Time-series analysis in operant research. Journal of Applied Behavior Analysis, 10, 151-166. Hartmann, D. et al (1980). Interrupted time-series analysis and its application to behavioral data. Journal of Applied Behavior Analysis, 13, 543-559. McDowall, D. et al (1980). Interrupted time series analysis. Sage. Hogenraad, R. et al (1997). The enemy within: autocorrelation bias in content analysis of narratives. Computers and the Humanities, 30, 433-439. [the abstract indicates that ITSACORR and other methods are recommended to deal with this problem] ------------------------------------------------------------------------ Stephen Black, Ph.D. tel: (819) 822-9600 ext 2470 Department of Psychology fax: (819) 822-9661 Bishop's University e-mail: [EMAIL PROTECTED] Lennoxville, QC J1M 1Z7 Canada Department web page at http://www.ubishops.ca/ccc/div/soc/psy ------------------------------------------------------------------------
