...Different in the beginning

[luciano basso]

Hi Team,

I need to analyze the effect of two independent variables (School and
Teacher) on a dependent variable (performance) on three points in
time. For that thought of using a mixed ANOVA. However, analyzing in
advance the values of the pre-test, I noticed that the performance is
different between the groups (both mean values (analyzed by ANOVA) and
variability (reviewed by Levine test). If the groups are different in
the beginning, I believe I can not use time as an independent factor,
so it was suggested to analyze the delta between the moments. I think
it is the best option, because it ignores the absolute values of each
group and focuses only change. What I am concerned is not consider
where the teacher and where he came out, which may itself be a factor
for a higher or lower delta.


If anyone can help you think about this design and the choice of how
to analyze it.

Thanks,

Lu
--
______________________
PhD. Luciano Basso
Motor Behavior Laboratory - LACOM
EEFE-USP

=====================
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]

I have a bunch of questions.  How many schools, and how many teachers?  Are teachers nested within schools?  Are students the unit of analysis, and if so, are they nested within teachers?  Is the variability related to the mean?  (If the SD is proportional to the mean, for example, log-transforming the data would stabilize the variances.)

luciano basso wrote

Hi Team,

I need to analyze the effect of two independent variables (School and
Teacher) on a dependent variable (performance) on three points in
time. For that thought of using a mixed ANOVA. However, analyzing in
advance the values of the pre-test, I noticed that the performance is
different between the groups (both mean values (analyzed by ANOVA) and
variability (reviewed by Levine test). If the groups are different in
the beginning, I believe I can not use time as an independent factor,
so it was suggested to analyze the delta between the moments. I think
it is the best option, because it ignores the absolute values of each
group and focuses only change. What I am concerned is not consider
where the teacher and where he came out, which may itself be a factor
for a higher or lower delta.


If anyone can help you think about this design and the choice of how
to analyze it.

Thanks,

Lu
--
______________________
PhD. Luciano Basso
Motor Behavior Laboratory - LACOM
EEFE-USP

=====================
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

[luciano basso]

Bruce and Team,

There are two schools (one public and one private) and two teachers
(one master and other no master) ... forming four independent groups,
each with 30 students. As the difference in variability, I'm talking
about the non-homogeneity of variance (observed by Levine's test).

If you need more information, I will gladly describe.

Thanks,

Lu

=====================
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

[luciano basso]

Hi Bruce and Team,

There are two schools (one public and one private) and two teachers
(one master and other no master) ... forming four independent groups,
each with 30 students. As the difference in variability, I'm talking
about the non-homogeneity of variance (observed by Levine's test).

If you need more information, I will gladly describe.

Thanks,

Luciano

=====================
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]

Just to be clear, are there two individuals (a master and a non-master teacher) who teach in *both* schools?  Or are there 4 individuals, each of whom teaches in only one school (with one master and one non-master in each school)?

Also, can you provide the N, mean and SD for each of the 4 groups?

Thanks.

luciano basso wrote

Hi Bruce and Team,

There are two schools (one public and one private) and two teachers
(one master and other no master) ... forming four independent groups,
each with 30 students. As the difference in variability, I'm talking
about the non-homogeneity of variance (observed by Levine's test).

If you need more information, I will gladly describe.

Thanks,

Luciano

=====================
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