Date: Mon, 14 Feb 2011 19:43:59 -0500 Mike Palij "SPSSX(r) Discussion" Mike Palij Re: Repeated Measures To: Gerónimo Maldonado text/plain; charset="iso-8859-1"

I've appended your first post below to point out why there is confusion over your design. Let me make the following points:

(1) You say you have 4 treatments and 4 concentrations which would suggest you have a 2-way completely within-subject design where every subject (hamster) is measured 16 times (i.e., 4 x 4 = 16 combinations or treatments and concentrations). So, there should 16 variables specified as containing the dep var (glucose level). So, we expect to Tx1Con1, Tx1Con2, Tx1Con3, Tx1Con4, Tx2Con1, ... Tx4Con4. This assuming that you used a full factorial design for the 4x4 within-subject factors.

(2) In your syntax below you have "week" instead of treatment combination and "Rx" which is specified as a between-subjects factor. Did you have different groups of subjects/hamster receive different treatments or was there one group that received the 16 combination I specify in (1) above? What exactly is "week" supposed to represent? Why is there only 10 weeks? You don't even mention week in your follow-up post.

-Mike Palij New York University mp26@nyu.edu

-------------------First Post-------------------- Hi Guys:

I'm really stumped trying to RUN a repeated -measures model in an experiment. Here's the descriptive:

Hamsters feeded on a daily basis for over a week. 4 different Rx, each one with 4 different concentration in mg. N for each Rx/concentration = 4 Outcome: Change in Glucose Level

Here's the syntax:

GLM week_1 week2 week_3 week_4 week_5_ week6 week_7 week_8 week_9 week_10 BY Rx /WSFACTOR = week 10 Polynomial /METHOD = SSTYPE(3) /POSTHOC = Rx ( TUKEY BONFERRONI T2 ) /PLOT = PROFILE( Rx week*Rx ) /EMMEANS = TABLES(Rx) COMPARE ADJ(BONFERRONI) /PRINT = DESCRIPTIVE OPOWER HOMOGENEITY /CRITERIA = ALPHA(.05) /WSDESIGN = week /DESIGN = Rx .

There's the problem: Output came empty, no univariate or multivariate results, no mauchly's test...

Any help would be appreciated!

--

----- Original Message ----- From: Gerónimo Maldonado To: SPSSX-L@LISTSERV.UGA.EDU Sent: Monday, February 14, 2011 6:29 PM Subject: Re: Repeated Measures

Hi Gene:

Sorry for the orthographic monstrosity... yes my design consists in 4 treatments:

* X * Y * Z * A

Each contains 4 concentration

* X (5mg) * X (10mg) * X (15mg) * X(20mg)... and so forth with each one of the treatments.

The outcome of the experiment is know which Rx @ what concentration is better in reducing C6H12O6 in mice

Thanks for any input!

===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@LISTSERV.UGA.EDU (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

Back to: Top of message | Previous page | Main SPSSX-L page