L7 - Parametric/non-parametric tests Flashcards
What is a paired samples T-test?
Parametric or non? report the BLANK of two BLANK samples
A parametric test in which we report the mean of two related samples.
What is a Wilcoxon signed rank test?
Parametric or non? report the BLANK
A non-parametric test in which we report the median.
When do we run a parametric test?
A type of test that is run when 4 specific parameters have been met.
What are the 4 parameters data must meet in order for a parametric test to be used?
distribution? BLANK of BLANK across conditions? number of participants are BLANK? what kind of data for DV?
Data are normally distributed.
There’s homogeneity of variance across conditions.
Roughly equal number of participants in each condition.
Continuous data (interval/ratio) for our Dependent Variable.
What is a non-parametric test and when would you run it?
A test that makes no assumptions about the data, can be run when one or more of the parameters have not been met for a parametric test.
What is the assumption of normality?
Meaning that the data is normally distributed.
What is homogeneity of variance?
checking whether the BLANK of the DV is BLANK across BLANK of the IV.
This is checking whether the variance of the DV is similar across all conditions of the IV.
What is the degree of freedom (df)?
max
The maxmimum number of values that are free to vary in a data set.
What are 3 ways can you check that data is normally distributed?
Run, check, plot
1) Run a statistical test of normality.
2) Check the Skewness and kurtosis values are within acceptable limits.
3) Plot a histogram, box or whisker graph and visually inspect for abnormality.
How do you work out if the data has homogeneity of variance?
divide BLANK by BLANK if less than BLANK it has homogeneity of variance
Check that the largest standard deviation is no more than 4 X bigger than the smallest standard deviation (divide the largest SD by smallest SD). If it is < 4 the data has homogeneity of variance.
