0
$\begingroup$

I'm measuring the influence of the use of male gaze and nudity in magazines to see if girls self-objectify or not. Now, I want to determine the influence of age.

Independent variables:   male gaze (yes / no) 
                         clothes (few clothes / lots of clothes)
Dependent variable:      Self-objectification score
Covariate:               Age

In the experiment, the participants were assigned to 4 conditions (male gaze - few clothes; male gaze - lots of clothes; no male gaze - lots of clothes; no male gaze - few clothes).

In my ANCOVA, I only want to determine the main effects and interaction effect between male gaze and clothing.

This was my output:

  • no significant effect for male gaze, neither for clothing or age
  • BUT significant effect for male gaze * clothing (interaction effect)

So, apparently, age had no significant influence on self-objectification, but the interaction effect between male gaze and clothing was significant (.010). What does this mean? That the participants who were assigned to this condition ALWAYS objectify? I have difficulties interpreting this, especially since the ANOVA showed that there was no significant interaction effect between male gaze and clothing...

$\endgroup$
1
  • $\begingroup$ I am not sure about your ANCOVA model, for example, is there any interaction term with age? You mean you did ANOVA, and the interaction between male gaze and clothing is insignificant, which is different from the ANCOVA model, am I right? It would be clear to write down both the ANOVA and ANCOVA models. $\endgroup$
    – Randel
    Commented Jul 21, 2013 at 16:42

1 Answer 1

2
$\begingroup$

Your interaction between male gaze and clothing means that their effects are not additive.

To better understand this, take a look at my example plot: the effect of male gaze with few clothes increases the score, but the effect of male gaze with lots of clothes decreases the score. The same happens with the effect of lots of clothes, it increases the score if male gaze is "no" but it decreases the score if male gaze is "yes". There's no such thing as a single effect for these variables since their effect depends on the state of the other variable.

Interaction plot

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.