|
Size: 1010
Comment:
|
← Revision 14 as of 2016-05-04 09:38:25 ⇥
Size: 1340
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 1: | Line 1: |
| You have done a repeated measures analysis in SPSS and find you have an interaction. In order to interpret this interaction you can analyse components of the interaction known as simple effects. Analysis of simple effects in a repeated measures design is only available using syntax. [http://www.utexas.edu/cc/faqs/stat/spss/spss50.html This syntax is described and illustrated]. | You have done a repeated measures analysis in SPSS and find you have an interaction. In order to interpret this interaction you can analyse components of the interaction known as simple effects. Analysis of simple effects in a repeated measures design is only available using syntax. [[http://www-01.ibm.com/support/docview.wss?uid=swg21475404|This syntax is described and illustrated here]] and is to be added on [[https://stat.utexas.edu/software-faqs/general | here.]] Field (2013, p.531) also illustrates this same approach (the /EMMEANS subcommand). |
| Line 12: | Line 12: |
| See also [[FAQ/Interaction| here.]] __References__ |
|
| Line 13: | Line 17: |
Field, A (2013) Discovering statistics using IBM SPSS Statistics. Fourth Edition. Sage:London. |
You have done a repeated measures analysis in SPSS and find you have an interaction. In order to interpret this interaction you can analyse components of the interaction known as simple effects. Analysis of simple effects in a repeated measures design is only available using syntax. This syntax is described and illustrated here and is to be added on here. Field (2013, p.531) also illustrates this same approach (the /EMMEANS subcommand).
Simple effects for a between subjects analysis may be performed using syntax with the MANOVA procedure. For example for comparing the two y means of factor b at level 1 of factor c we run
MANOVA Y BY b(1,2) c(1,3) /DESIGN b WITHIN c(1) VS WITHIN.
The WITHIN term above is the error term from fitting the full factorial model (b, c and the interaction b x c) and is recommended as the error term in simple effects involving only between subjects factors (Boniface, 1995, p.155).
See also here.
References
Boniface, DR (1995) Experiment design and statistical methods for behavioural and social research. Chapman and Hall:London.
Field, A (2013) Discovering statistics using IBM SPSS Statistics. Fourth Edition. Sage:London.