week 5 Flashcards
parametric tests
compare means of groups
assumes that data are numeric and that data is random with a gaussian distribution of variables
one sample t test
comparing group mean with a reference value
independent sample t test
comparing means of 2 groups
paired t test
all participants measures pre and post intervention, compare means
one way anova
compare means of >2 groups
p value, post hoc
repeated measures anova
all participants tested under >2 conditions/times (ie ROM at 0,3,6 weeks)
2 way anova
Main effects: A,B
Interaction: AxB
difference in means according to 2 or more factors
or difference in means among 2 or more groups with repeated measures (interaction btw time and factor)
interaction
differences in the effects of one variable according to the level of another
non parametric tests
parametric assumptions not met, non gaussian dist
parametric independent test
two sample t test
parametric paired
paired t test
non parametric independent
mann whitney U
nonparametric paired
wilcoxin signed rank
assessment of linear correlation
p value: is there a correlation?
r value: how strong?
.75= excellent