[SPSS - Binary logistic test - Interaction (moderator) ]

[Kelvin86]

DV: dichotomous (0 = show, 1 = no-show)

IV (covariates):

[translation]
Geslacht = Gender
dvd_week = Day of the week
[/translation]

Gender (0 = females, 1 = males) --> Indicator(first)
dvd_week = (1 = monday, 2 = tuesday, ... , 7 = sunday) --> indicator(first)

Interaction (moderator):
dvd_week * geslacht
------------------------------------------------------------------------------------------------

Method: Enter

------------------------------------------------------------------------------------------------

As you can see from the results (Variables in the Equation).
1st question: Is the interaction significant or not? .. well obviously it's not, but it's 'children' are. So what do I write in my thesis? Is the interaction (moderator) between Gender and dvd_week significant or not?????? Is there a statistical way of writing this?

2nd question: What is the difference of the star ( * ) and by? .. what is SPSS telling me here?

 

[Kelvin86]

Yes, it's a very large sample. About 900 thousands users. I already asked my supervisor, but he's too busy, and can only answer until next week monday. Can you help me, so I can progress?

I already searched through the forums, I found no other same as this.

 

[victorxstc]

Of course!

First question: I couldn't find that specific interaction in the PDF file. But look at its Sig. (which is P value). If it was less than 0.05, it is significant (u already know I think). If so, write "the effect of DVD_week (on your dependent variable, whatever it is) differed for each gender". Otherwise, write "the effect of dvd_week on dependent variable was similar for men and women".

question 2: I think the asterisk is used for the factors as whole things (dvd*gender), and "by" is being used for each level of each factor (dvd1 by gender1, dvd2 by gender1, etc.). I think both mean "interaction". Please give more details on your study variables and your design.

 

[Kelvin86]

Hey victorxstc,

Yes the dvd_week * gender interaction is p = 0.098 (middle of the page), so this is not significant, but it's other categorical members (look below in the pdf) some of them is significant and others not. .. the question is why is dvd_week * gender interaction not significant, but the others are significant???