Statistical Tests Flashcards
Statistical tests for comparing group means
Independent t tests (Mann Whitney u non parametric alt
Paired t test ( Wilcoxon)
One way ANOVA (Kruskal Wallis)
Repeated measures ANOVA
What statistical test do you use for looking at relationships between variables
Pearson correlation (spearman non paramteric)
Simple Linear regression
Multiple linear regression
What would you use as an alternative for a repreated measures ANOVA
Friedmans test
Compares the means of two independent groups with one categorical and one continuous variable
Independent t test, you want to know if group a and group b are different in their averages eg comparing bmi between men and women
If mauchlys test is significant (p< .05) what does that mean?
The assumption of sphericity is violated
What do you do if the assumptions of sphericity is not met
Apply corrections like greenhouse geisser
When greenhouse geisser correction is applied, how do the degrees of freedom look
The correction reduces degrees of freedoms to decimals which makes test more conservative
What type of data is required for repeated measures ANOVA
One continuous variable across 3 or more related conditions
You run a repeated measures aniva and get p = 0.058 after greenhouse geisser correction, what do you conclude?
The result is borderline but not significant
What does the greenhouse geissee row in the spee outfput adjust
The degrees of freedom
You want to compare the average test scores between two independent groups, what test do you use
Independent t test, it compared the means between two separate groups
Whats the difference between and independent t test and a pair t test
Independent t test compares the means between two separate groups
Paired t test is for the same group measured twice
Assumptions of the independent t test
Normal distribution of The dependent variable
Homogeneity of the variance (levenes test)
Independence of observations
Levenes test is significant what does that mean
Equal variance are not assumed use welchs test
Two continuous variables you’re checking for a relationship
Pearson correlation
Pearson correlation assumptions
Borth variables are continuous
Linearity
Normality
No significant outliers
What test do you use to compare 3+ independent groups
One way ANOVA
One way ANOVA assumptions and what to do if assumptions aren’t met
Normal distributions within each group
Homogeneity
Independent observations
Kruskal Wallis is. Non parametric alternative
What test do you use to compare the means across 3 time points
Repeated measures ANOVA
repeated measures ANOVA assumptions
Sphericity, normality
kruskall wallis uses
medians and means
medians are used in a kruskal wallis when _
the distributions of the groups are the same
means are used in kruskal wallis when __
distributions of groups are not the same shape
kruskal wallis tells you
if there are differences between groups but not where they are
two tailed alternative hypothesis
true mean does not equal hypothesized value, looking at IF there is a difference, not which direction
in a one way ANOVA independent variables are referred to as
grouping factors, which are split into levels
if 95% confidence interval contains 0 then groups are
not different
Whats the point of a post hoc test
To see where the differences are in an ANOVA
Coeeffience go determination the closed r2 is the 0 the x has to do with ya variability
Less
This score is used when you have the population standard deviation and you’re comparing a single score
- tells you how many standard deviations a value is from the mean
Normal distribution
Z score
Used when you don’t gs e population standard deviation and you’re working with a small sample less than 30
Used fish n t tests to compare means between groups or before and after
Wider than normal distribution
T score
Used when comparing variances or more than 2 groups
Used in ANOVA to test iif at least one group mean is different
Positively skewed
F score
