Research methods: Irene. 7-3-16
Post hoc and planned comparisons in exam.
With planned comparisons: Check whether sig value in output table on spss says 2 tailed, if
it does have two tailed but hypothesis was directional then can divide the sig value by two
and use that value, making it more significant.
Uni variate means one dependent variable.
Levene needs to be not significant – needs to be greater than 0.05
Is using spss for post hoc and bonferonni is chosen then don't have to do any dividing by
groups etc as spss accounts for that with bonferonni correction.??
In exam: shown interaction graph and will ask if it looks significant, then ask what type of
tests will be used between post hoc and planned comparisons. Planned comparison if
hypothesis was directional.
Post hoc and planned comparisons in exam.
With planned comparisons: Check whether sig value in output table on spss says 2 tailed, if
it does have two tailed but hypothesis was directional then can divide the sig value by two
and use that value, making it more significant.
Uni variate means one dependent variable.
Levene needs to be not significant – needs to be greater than 0.05
Is using spss for post hoc and bonferonni is chosen then don't have to do any dividing by
groups etc as spss accounts for that with bonferonni correction.??
In exam: shown interaction graph and will ask if it looks significant, then ask what type of
tests will be used between post hoc and planned comparisons. Planned comparison if
hypothesis was directional.
