SPSS Statistics

SPSS Statistics

Your hub for statistical analysis, data management, and data documentation. Connect, learn, and share with your peers! 


#Analytics
#SPSSStatistics
#Analyticstools
 View Only

  • 1.  Significant differences in means with one sample group

    Posted 05/09/22 09:13 AM

    Hi guys,

    I am somewhat of a newbie when it comes to running analyses in SPSS. My query is about the following: I did a survey on product involvement. I asked participants 4 questions about their involvement in 6 products. Im trying to figure out how i can analyse the significant differences in the means per product and question, however I am not sure on what test to use and how to run it SPSS. I have tried doing the MANOVA, however I don't know what to use as a factor. 

    Any suggestions? Let me know!



    ------------------------------
    Sarah Mohamed
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Significant differences in means with one sample group

    Posted 05/09/22 09:56 AM
    Edited by System Admin 01/20/23 04:35 PM
    Hi. A factor would be some variable that divides the people you measured into 2 or more groups, like a control and experimental group. You don't *NEED* such a variable; you can still run a model like:

    GLM question1 to question4

    or 

    GLM product1 to product6

    which are multivariate analyses.

    It's hard to give you a straight answer without knowing more about your data and exact research questions.

    ------------------------------
    Rick Marcantonio
    Quality Assurance
    IBM
    ------------------------------



  • 3.  RE: Significant differences in means with one sample group

    Posted 05/09/22 07:46 PM
    Edited by System Admin 01/20/23 04:14 PM
    Hi Rick,

    Thank you for your reply! I did not conduct an experiment. I did ask the general questions about demoghraphics, however, I want to analyze the differences in the means of the variables (question outcomes). The variables/questions are 4 Likert scale statements about the involvement of consumers during and after the purchase process of products (e.g. I would be interested in reading customer reviews about this product). Someone suggested that perhaps an easier route would be a factor analysis. I've done that: I've put the 24 (6 products x 4 questions) variables in the analysis, but I am not sure on how to interpret the output; how do the factor scores in the rotated component matrix show the significant differences (if thats even the case)?


    ------------------------------
    Sarah Mohamed
    ------------------------------

    Original Message


  • 4.  RE: Significant differences in means with one sample group

    Posted 05/09/22 08:51 PM
    Edited by System Admin 01/20/23 04:45 PM
    If you really want to look at the differences between sample means, then you should consider repeated-measures analysis of variance. 

    For just two products, that would look something like this:

    GLM prod1_rating1 prod1_rating2 prod1_rating3 prod1_rating4
    prod2_rating1 prod2_rating2 prod2_rating3 prod2_rating4
 /WSFACTOR=rating 4 Polynomial product 2 Polynomial
 /METHOD=SSTYPE(3)
 /CRITERIA=ALPHA(.05)
 /WSDESIGN=rating product rating*product.
​

    I know you have 6 products. It would just be more variables and "product 6", not "product 2" in /WSFACTOR.


    I'm not sure factor analysis is what you really want. Perhaps whoever suggested that was thinking of  correspondence analysis, since you are comparing several different products along some dimensions. That is more of a contingency-table type of approach.

    In any case, you should work personally with a statistician who can look at your data with you and help you choose the right analysis, the set-up for it, and exactly what the results do (and do NOT) mean.

    ------------------------------
    Rick Marcantonio
    Quality Assurance
    IBM
    ------------------------------

    Original Message


Related Content