Math Forum / Math Software / SPSS / August 2006

Thanks in advance for your advice.
I'm studying a disease and I have prevalence values for each of four
groups of individuals.  The characteristic that designates each group
is an ordinal variable.  Here is an example:

This is probably a stupidly simple question, but the documentation
doesn't help, no one else in the office knows, and I can't find
anything on Google.
I have a set of string variables with values A-D and 1-4 in each

I am using SPSS v10 so far only manually. However, I want to automate
my tasks. I now quite a bit about using VBA in Excel and am wondering
if there is kind of a macro recorder in SPSS as well to quickly get the
code pieces I need and that can be adapted then.

please can someone help me with the following.  It's been a few years
since I used SPSS and now I find myself having to get back into it
rather quickly.  I am using version 12 in a Windows environment. The
SPSS files I am supplied with ought not to contain any variables with a

How can I do this the fastest in SPSS (v13)?:
There are 20 variables.
There are 2000 rows.
There are 10 countries; (on average) 200 rows per country.

I've been trying to compute a new variable which will have 4 categories
(both parents lean, both parents obese, father lean/mother obese,
father obese/mother lean) from fathers & mothers BMI class
(1=underweight, 2=normal weight, 3=overweight, 4=obese)

I am quite new with SPSS and have the following problem:
I have about 100 Excel spreadsheets with data. (Each is one patient).
Every spreadsheet has the variables x and y. (always as pairs, each in
one column).

I really need some help with the stats for my dissertation.  I
basically ran an audio personality test on 15 people. Half the
questions were asked in a human voice and half in a
computer-synthesised voice. The personality test measures 32

I have a dataset which records ECG changes in variables v1 to v6 as
positive and negative numbers as well as zeros (if there is no change).
I want to be able to
1.  Add the positive numbers from v1 to v6 to get a cumulative "score",

I'm struggling with the following: for a certain variable i want to
replace the missings with the value from the previous case. data is
organised like this:
1    and I want...        1

Hey everybody,
in case I like to compare several means between groups on a intervall
scaled variable, I use a ANOVA.
What can I do, in case my dependent variable is scaled ordinally? So to

First of all, sorry for my english but its not my mother tongue. I have a small problems with
the frequencies or the crosstabs analysis. To explain it correctly I will give an example.
I have for items that I want the frequencies :

My data file has over 1000 variables and I need to sum up some values
of some cases (more than one case for the same person). I managed to do
this with the aggregate command, but my syntax now looks like this:
AGGREGATE

I would like to determine if a variable differs between two groups. The
independent variable is group membership yes/no. The dependent variable is
present yes/no. Therefore, both variables are dummy's.
It was suggested to me to use the z-score. How can I do this analysis in

I'm wondering if anyone could tell me about syntax  of post
hoc(Sheffe)?
I'm running 2-way ANCOVA and I wanted to controll for homosexuality in
my data as I've got quite few gay population which is not my main