On Mon, 26 May 2003, SORGER wrote: > i have data with different factors: > cultivation (waterstress/no), chambers (fumigation/no), sorte (5/6); > in one column i have put for cultivation 1/2, for chamber 3/4 and > for sorte 5/6. yield is the dependent factor.
Do you mean you have a column of only one digit that has six possible values (1,2,3,4,5,6)? How do you distinguish between a case that had waterstress AND fumigation, and another case that had waterstress but NO fumigation? And so on. It would probably help if you showed an excerpt from your data file, along with information showing what variables (factors) are where in that excerpt. > now i want to evaluate if the difference between the 2 cultivations > and fumigations is significant. Which difference? I see at least two differences: one between the two levels of cultivation, and another between the two levels of fumigation. There ought also to be a difference, possibly more complicated, that expresses the interaction between cultivation and fumigation. > does someone know how i can evaluate these data with statistica? > do i have to separate the columns? I'm not quite sure what you mean by "separate the columns" (since that depends on what exactly is IN the columns!), but I rather suspect the answer is "Yes". > and second: i have fixed conductance-datas and many different > environmental influences. i have to evaluate these data with > multiple regression. can someone give me advices? Maybe, if we knew more about the problem. "Many different environmental influences" suggests to me an observational study rather than an experimental study; is that what you meant to imply? And for what purpose(s) do you "have to evaluate these data", or with what kind of result in mind? ----------------------------------------------------------------------- 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/ . =================================================================
