1 way independent non-parametric ANOVA

Question 1

how are the scores ranked?

Answer 1

all scores are ordered regardless of the group they were in > scores that are the same are ranked in order > these ranks are added together and divided by number of scores to get a mean value of the ranks > they are then reordered in ppt order for each condition they were originally in > then calculate mean rank of each condition

Question 2

why do we use ranked data instead of using raw scores?

Answer 2

it evens out the spread of data and differences. this helps to stop having a key in the data that is created by the outlier

Question 3

what are the outputs from the Kruskal wallis test?

Answer 3

mean rank, Kruskal Wallis H statistic, df, asymp Sig (p value)

Question 4

how do we know if there is a significant difference?

Answer 4

the p value is less than 0.5

Question 5

if there is a significant difference between groups how do we find out where the difference lies?

Answer 5

conduct post hoc tests

Question 6

for non parametric independent ANOVA what non parametric version of the independent t test is carried out?

Answer 6

Mann-Whitney

Question 7

how would the output be written for the Kruskall wallis test?

Answer 7

H(df)= [H value], p = [p value]

Question 8

what are the characteristics for Mann Whitney post hoc tests to be carried out?

Answer 8

continuous DV + categorical IV > between ppts > violation of assumptions so can’t carry out independent t test > more than 2 groups > Kruskall wallis is significant

Question 9

what alpha value is used for mann whitney tests to check significance?

Answer 9

bonferroni corrected alpha of 0.016

Question 10

what needs to be reported for the write up of Mann-Whitney post hoc tests?

Answer 10

median values, Mann-Whitney U value, p value

Question 11

list the steps to run an independent non parametric ANOVA

Answer 11

ANALYSE > NON PARAMETRIC TESTS > LEGACY DIALOGS > K INDEPENDENT SAMPLES. move dependent variable into ‘test variable’ box > click Kruskall Wallis H > click define range > enter lowest and highest value given to a group > continue

Question 12

list the steps to run post hoc Mann Whitney tests if the non parametric ANOVA is significant

Answer 12

ANALYSE > NON PARAMETRIC TESTS > LEGACY DIALOGS > 2 INDEPENDENT SAMPLES. move dependent variable into ‘test variable list’ box and move factor into ‘grouping variable’ box > click on Mann Whitney U