Math Forum / Math Software / SPSS / February 2007

I work on a prediction model for bank takeovers. I use a logistic
regression with a dependent variable with value 0 (nontarget in
2003-2004) or 1 (target in 2003-2004) and a number of metric
independent variables (financial characterisics at 1-1-2003).

Is it possible to pass values into a syntax script where script
variables are substituted with those values?
For example:
spss script.sps 11 "a string value" 4

I've been wrestling with this problems for a long time and am hoping
others might have some insight.
Let's say I run a factorial anova and get an interaction and would
like to graph the emmeans. I can right click on the pivot table (or

The following is the simple scatterplot example from the syntax
reference guide (v15):
GGRAPH
/GRAPHDATASET NAME="graphdataset" VARIABLES=salbegin salary

I am trying to enclose a GGRAPH/GPL command in a macro.  For GGRAPH and
GPL, it is essential that one use double-quotes.  However, even though I
use double-quotes in my macro definition, they become single-quotes when
the macro is expanded.  E.g., here is a macro definition:

Does anybody happen to know how I can (easily) calculate d-prime in SPSS?
Regards,
Martijn

I'm writing a script where I loop over the variables of a data set.
However, I want to treat multiple response sets as an entity.
Is there a way to get a collection of the Multiple Response Sets
defined and then investigate individual sets in detail; e.g. 'Number

I typically employ Cohen's Kappa (calculated with SPSS) to measure
inter-rater agreement on nominal variables, but this time I need to
measure rater agreement among three raters.  Is this possible in
SPSS?  While looking for information, I came across the text below but

My daughter is using SPSS v12 to analyse medical data.   She is using cross
tabulation and calculating chi sqaureds.  As some of the cells contain less
than 5 the Chi squared test is not valid.  As most of her tables are not 2x
2 but 2x3, the Fisher's exact test will not work.  I ...

My examination was preformed with help of questioner with 122
questions, and the data I’ve processed with statistical program SPSS
13.
First I’ve found the percentage(frequency) and arithmetic mean  which

I need help with some complex analysis.
I will show you only two important
fields as a simple example:
CreditType               ID

In complex medical data, each row represents each medication for one
visit for one patient and its adverse event from each medication.  The
data looks like this.
ID    Visit    Medication   AdverseEvent1

I use SPSS for sime years, but now I really do have a problem. We are
testing politicians and candidates with different questions that are
built on two dimensions that are interesting for us. I would like to
either/or analyse every case as

I have set of indept variables (0 - No, 1 - Yes basis) and dependnet
variable in 9 point scale.
In this situation which correlation analysis I should consider.
Thanks,

Is there a way I can produce output that lists all of the unique
values in a group of variables: For instance:
Variable 1 values: 0, 0, 1, 1, 1, 2, 9
Variable 2 values, 0, 1, 1, 2