On 18 Jun 2003 15:08:24 -0700, [EMAIL PROTECTED] (Dianne Worth) wrote: [snip, some] > > My next question is this: > Rather than multiple regression, I want to use structural equation > modeling. The sample is < 250, which I've been told is not large > enough for LISREL. The alternative software uses partial least
The N needed would depend on the number of predictors and the size of the correlations. With single items as predictors, the correlations will be rather small, so you surely can't model a whole lot of them with N=250. But that also depends on the shape of the model. Do the questions fall into three or four pre-formed, hypothesized factors? > squares for analysis, but will not accept negative numbers. So I > converted as before and then multiplied by -1. Does this make sense? The program won't accept negative numbers among the data? Is that a really cheap, low-end, home-made program? Even my own homemade programs always accept negatives, at least if negative is valid. - I conclude that you are probably in trouble. Either the program is inferior; or you are misreading, and trying to feed raw data where it wants some processed, positive number. > I am using Ajzen's Theory of Planned Behavior and will be glad to > supply any additional information, just don't want to go into > information overload. > I had not heard of that theory. Google found a surprising number of hits when I misspelled the name as Azjen, and enough extra, real hits that it offered the correct spelling. Here seems to be a nice summary, as part of an applied document by someone's Department of Transportation (the .dot. in the name). http://www.fhwa.dot.gov/tfhrc/safety/pubs/96143/appa/body_appa_07.html > [ snip, earlier post and subsequent comments by dfb.] -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html "Taxes are the price we pay for civilization." Justice Holmes. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
