help for monava analysis

Dear Proffessor
Dear Authorized
I made manova analysis with two independent variables (that are gender and
grade level) and ten dependent variables that have got between .30 and .60
correlations.

The result of analysis is that manova found a significant interaction but
multiple anova didn’t find any significant interaction effect after
bonferroni correction (alpha/10).

In this case, what Can I do? I read your using multivariate statistics,
but it is explained the reverse of my problem that manova is non
significant but multiple anova is siginificant.

Thank you very much


Dr. Arif ÖZER

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Salbod]

Dear Arif: The bonferroni may be too conservative (p = .005). You should consider using modified a bonferroni, such as the Holmes-Bonferroni procedure. I hope this  helps. --Steve

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of [hidden email]
Sent: Monday, November 08, 2010 9:45 AM
To: [hidden email]
Subject: help for monava analysis

Dear Proffessor
Dear Authorized
I made manova analysis with two independent variables (that are gender and grade level) and ten dependent variables that have got between .30 and .60 correlations.

The result of analysis is that manova found a significant interaction but multiple anova didn’t find any significant interaction effect after bonferroni correction (alpha/10).

In this case, what Can I do? I read your using multivariate statistics, but it is explained the reverse of my problem that manova is non significant but multiple anova is siginificant.

Thank you very much


Dr. Arif ÖZER

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Bruce Weaver]

Do you have a multivariate question, or a series of univariate questions?  If your questions are univariate, there is no real need for a preliminary MANOVA.  If your question is multivariate, the univariate ANOVAs are probably not really needed.  See Huberty & Morris (1989) for discussion of this.

    http://management.uta.edu/casper/multistat/huberty%20&%20morris%201989%20-%20manova%20vs.%20anova.pdf

HTH.

[e-mail] wrote

Dear Proffessor
Dear Authorized
I made manova analysis with two independent variables (that are gender and
grade level) and ten dependent variables that have got between .30 and .60
correlations.

The result of analysis is that manova found a significant interaction but
multiple anova didn’t find any significant interaction effect after
bonferroni correction (alpha/10).

In this case, what Can I do? I read your using multivariate statistics,
but it is explained the reverse of my problem that manova is non
significant but multiple anova is siginificant.

Thank you very much


Dr. Arif ÖZER

=====================
To manage your subscription to SPSSX-L, send a message to
[e-mail] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD