Math Forum / Math Software / SPSS / September 2006

Can someone throw some light on the Durbin-Watson test for autocorrelation?
What is the decision rule if result is less than 1. Result is 2 or more than
2? What's the significance of the test? Do you always run this test before
doing forecasting?

Dear SPSS Support Group:
I have four columns of data under the variable names:
DrugName, Volume, CountyCode, and Price.
To generate my report I'm selecting:

I am using 'Moderated Multiple Regression(MMR)' to test the moderating
effect of social support on the relationship say work-family conflict
and job satisfaction. I tried but failed to procure an interaction
graph from SPSS (11).

Does anyone know if it's possible/how to add the current date at the
end of a file name,
I have syntax that I run repeatedly that is saving files as a specific
name. Each time I run it I'd like to give it a different name without

I want to conduct a categorical principal components analysis using
SPSS 12. My aim is to see which dimensions I can use to describe
reactions to health risk feedback (reactions being partly measured by
questionnaire and partly derived from content analysis, hence partly

I am trying to advise a colleague of mine on how to analyse a
cross-over clinical trial using GLM in SPSS. Usually, I would perform
the analysis in SAS. I am having trouble in SPSS trying to specify the
correct model.

I have an data set which has both numeric and string variables. I need
to arrange in such a way that all the numeric variables comes first,
then sting variables. Any way to do this?
Advance thanks

I want to randomly generate data to simulate reponses to a multi-item
questionnaire - e.g. discrete values (1,2,..7) on a dozen separate
items. I'd like the responses to follow a beta distribution, and I'd
like to be able to specify the correlations between items.

I am running a multivariate repeated measures ANOVA, with the dependent
variable measured at five time points. I have three between-subjects
factors and two covariates. My difficulty is in understanding exactly
what the estimated marginal means are adjusted for. In models WITHOUT

I want to recode all the numeric variables in a database that contains
lot of variables. However some of them are string or date variables.
This recode does not work because there are not all numeric variables
Ex.  Recode v1 to v800 (sysmis=999).

SPSS13
Situation
Population: 10.0000
Stratification: Gender

I am using SPSS 14 on a Windows XP Home machine. From a internet survey I
got a data matrix (csv file) with 53 variables and 16 cases.
I imported the data into SPSS and it worked pretty well, but if I try any of
the analysis functions I get the error: "All variables in the working ...

Hi friends,
Firstly a brief introduction. I have been using different versions of
SPSS from 1975 to 1996. First versions were various mainframe versions
on different mainframe platforms, only command language oriented. Later

I am trying to analyse the results of a campaign of activity to a group
of customers.
We have a given group of customers (around 130,000) that over 2 years,
may have received a sales call although not all customers will have.

I have a MANCOVA with:
- a dichotomous variable (the dependent variable)
- a categorical variable
- a continuous control variable