Math Forum / Math Software / SPSS / June 2005

I try to establish a typology of countries, some of which have a number
of missing values. Therefore I use discriminant analysis to find for
each country the type it is most close. However, I dont find in the SPSS
10 Base Guide the appropriate procedure which classifies cases ...

I am trying to understand the meaning of the contrast coefficients in
logistic regression. For an analysis to test whether the categories of an
independent variable are truly ordinal, I invoked the repeated contrast
option for categorical variables.

I hope someone can help me with this data transformation problem,
because I'm working with a big enough dataset that I really don't want
to do this by hand (I have neither the time nor the finances).  I am
using SPSS 13.  I have a dataset that looks like this (n=231):

I try to compute the median of a variable called income. As you cant do
it with a Compute command I use an AGGREGATE command but this creates an
error, and I dont understand where the error is.
The working file is sorted by the variable COUNTRY.

I once copied my old SPSS Tables manual, with excellent examples, and
how to write code. Not having it used for a long time I can't find it
any longer. Does someone know whether I can find it on the net ?
Cheers

I found an old question about this but unfortunately no replies. I just
want to have a series of barcharts/dotplots together (like you can
easily in minitab and splus), but can't seem to do it in spss. The
panel function appears to be what I want but splits up one variable by

I am trying to produce a graph showing categorical changes in a variable at
3 time periods for 100 individuals.
I want to display all this on one graph - is that possible?
Regards

I tried to transfer data from an excel-sheet to SPSS. But
the results were not satisfying: it was unpredictable, which
variable was transferred as string-variable and which one as
number. I even added one line of data (as the second row),

I did a linear mixed model analysis with the result below.
Type III Tests of Fixed Effects
Source     Num df Denom df  F        Sig.
Intercept  1      81.041    16.817   .000

We're interested in seting up a credit card programme in the EU where
credit management is built on behavioral statistics rather than
personal opinion or plain credit score.
This is an ununsual approach in Europe outside the UK.

I need to convert the date field from '01jan1989' to spss date and
caculate age. How do I do it? Thanks.
Max

The problem is to select cases from about 150 pairs of cases like
this:
For all cases where var 1 in one case = var 2 in another case, then of
these 2 cases delete the case for which the value of var 3 is lowest.

I have 5000 compositions rated by 2 raters.
I have entered the data in SPSS in a table that looks as follows:
Student    Compositiongrade1    CompositionGrade2     Correction    Rater
1234                   15                                  12

I've been exporting some data to Excel (~30K rows, ~50 variables).
I've been using the SAVE TRANSLATE command syntax, as generated by using
the menu:  File => Save As.
I've had two problems - first, the export rate is only ~6 cases/second.

ng,
I'm trying to save some excel files. I use the same *.sav throughout
the procedure, but with filter and /unselected=delete I thought it
should be quite easy. But no...