Run a series of separate Anovas, one for each importance statement. (Analyse - General Linear Model - Univariate). Under Options, click Estimates of Effect Size. Partial eta-squared will be printed next to the significance level. When there is only one factor in a given analysis, partial eta-squared and eta-quared are the same thing: an estimate of the proportion of variance in the DV that is accounted for by that factor.

You then rank the attributes in order of eta-squared.

A couple of caveats:

- This will not tell you whether any differences in eta-squared are
*significantly*different from each other - i.e. two attributes may be equally diagnostic, but one might just seem bigger due to sampling error. - The estimates may be inaccurate (probably too low) if assumptions are not met, e.g. if there are non-normal distributions (which seems particularly likely if they are attribute statements). And this inaccuracy may well be greater for some attributes than others, putting those in the wrong place in order.

Mike Griffiths

Date: Fri, 2 Jul 2010 11:55:25 -0700

From: johnwatson_1981@yahoo.com

Subject: Require urgent help on ANOVA

To: SPSSX-L@LISTSERV.UGA.EDU

Hi All,
I need urgent help on the following situation. Would really appreciate your help.
I need to use ANOVA on the following variables:
1. Whether purchased product A - Yes or No (consider this factor variable)
2. Series of ratings on attribute importance statements
Objective: I need to run ANOVA to find out which attributes' means are significantly different between those purchased vs. non-purchasers. Once I find the attributes that have higher mean on those who purchased, I need to find out which attribute is more predictive or influencing in likely to purchase the product. In nutshell, I need to rank order the attributes by their ability to impact purchasers vs. non-purchasers.
Help: Can ANOVA analysis help to accomplish aforementioned objective? If yes, how?
I am totally lost and I need to get this out asap. I would really really appreciate any help you can provide. |

Get a new e-mail account with Hotmail - Free. Sign-up now.