Date: Tue, 27 Aug 2002 11:16:57 -0400 "Zdaniuk, Bozena" "SPSSX(r) Discussion" "Zdaniuk, Bozena" standard deviation from standard error text/plain; charset="iso-8859-1"

Hello, everybody.

I'm trying to compute an effect size (d) from post-treatment data for 2 groups: Mean for group 1 - Mean for group 2, divided by the pooled standard deviation.

I'm pretty confident about computing the pooled standard deviation, but the d is still not coming our right (it's greater than 1!). I think this is because the data that I am working with are unadjusted means whereas the RM analysis included 2 covariates. So, I asked SPSS for adjusted means. Now my problem is that they print out a standard error with the adjusted means, instead of a standard deviation. Do you know how to calculate a standard deviation from a standard error? I need the SD in order to calculate the pooled standard deviation.

Please let me know if you can help.

thanks a lot. bozena zdaniuk

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] Sent: None To: SPSSX-L@LISTSERV.UGA.EDU Subject: Excel to SPSS-num-string

Neil, SPSS establishes the nature of the variables (string or numeric) by the value they have for the first case (first data row) in the spreadsheet. If some of your variables have blanks -or other nonnumeric character- for the first case, they are regarded as strings. Make sure there are no blanks in the first row of data (if necessary, include a phony first case with data of the correct nature, to be deleted afterwards).

Hector Maletta Universidad del Salvador Buenos Aires, Argentina Neil Cobb wrote: > > In the process of opening an excel worksheet from an excel workbook (Excel > 2000) in SPSS 10.0.5, sometimes many variables are read as strings and > not numeric values, and it a hassle to convert each variable from string to > numeric. Some variables are read as numeric, but I cannot figure out how > they are different from variables that are read as strings. Is there > either a global command in SPSS to change sets of variables from string to > numeric, or a way to prevent SPSS from reading them as string variables in > the first place? > > Thanks in advance > ********************************* > Neil S. Cobb > Department of Biological Sciences > Northern Arizona University > Flagstaff, AZ 86011 > Fax: 520-523-7500 > Phone:(w) 520-523-5528 > (h) 520-214-6237 > e-mail: neil.cobb@nau.edu > http://pinyon.bio.nau.edu > **********************************

Back to: Top of message | Previous page | Main SPSSX-L page