Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[Q] How to interpret MANOVA result

0 views
Skip to first unread message

Young-Jin Lee

unread,
Jan 22, 2003, 11:10:08 PM1/22/03
to
I found my previous post is difficult to read. So I try to make it easy to
read again.

I am looking for a help in interpreting a SPSS MANOVA results. I looked up a
manual and search the Web, but I was not able to find the answer.

I did a MANOVA (Analyze/General Linear Model/Multivariate...) with 14
dependent variables and 2 group variables. There were 14 subjects in each
group.

From the resources I found on the Web, I found that (1) the following table
shows that there is a statistically significant difference between groups,
and (2) I need to focus on "Group Effect". My first question is what is
"Intercept" effect? And from the name of "Group Effect" it seems to
represent differences caused by each group. Is it a correct interpretation
for "Group Effect"?

The second question is what are df.s of F statistics? Since the group effect
is what I need to pay attention to and assuming that I am using Wilk's
Lambda, then what are degrees of freedom for the F statistics (F = 16.007)?
I also want to know what are Error df and Hypothesis df.

Multivariate Tests

Effect Value F Hypothesis df Error df Sig.
...
Intercept
Pillai's Trace .986 77.460 14.000 15.000 .000
Wilks' .014 77.460 14.000 15.000 .000
Lambda
....

Group
Pillai's Trace .937 16.007 14.000 15.000 .000
Wilks' .063 16.007 14.000 15.000 .000
Lambda
...

I also have similar questions on the table named of between-subject effects.
Due
to the size of this table I cannot copy-and-paste the table, but it looks
like the below table. (DV = dependent variable)

TypeIII SS df MS F Sig
Corrected Model DV1
....
DV14

Intercept DV1
...
DV14

Group DV1
...
DV14

Error DV1
...
DV14

Total DV1
....
DV14

Corrected Total DV1
...
DV14

First of all, what are "Corrected Model, Intercept, Group, Error, Total,
Corrected Total"? I think this question is probably related to the one of
the previous questions.

Second, I got 1 df for corrected model, intercept, and group; I got 28 df
for error; I got 30 df for total; I got 29 df for corrected total. Can
anyone explain where these dfs from?

Finally, which F statistics should I report and what are dfs?

Sorry for the long question, but I really need to figure this out.

I hope to get answers.

0 new messages