a test of difference: Whitney and Wilcoxon

Question 1

why would we choose Mann -

- a worked example

Answer 1

e.g.
we are looking for a DIFFERENCE between two groups of employers based on their rating of whether a candidate with schizophrenia was suitable for a job interview

there are TWO INDEPENDENT groups of employers this means that the design is UNRELATED

Finally, the level of measurement is ordinal data as it is based on scores on an “unsafe” scale which are converted to ranks for the purpose of the test

Question 2

what are the hypothesis

Answer 2

ALTERNATE HYPOTHESIS: there is a difference in ratings for suitability for an interview based on whether a job applicant is described as having been diagnosed with schizophrenia (Group A) or not (Group B)
This would be non - directional two- tailed

NULL HYPOTHESIS: there is no difference in ratings for sustainability for an interview based on whether a job applicant is described as having been diagnosed with schizophrenia (Group A) or not (Group B)
This would be non - directional two- tailed
t

Question 3

what are the steps in a Mann - Whitney test

Answer 3

Step 1: the table of ranks

Step 2: working out the value of u

Step 3: the calculated and critical values

Question 4

how do we know whether the data is significant or not

Answer 4

if the calculated value is less than the critical value, the results are significant

this means we can reject the null hypothesis and accept then alternate hypothesis

Question 5

why should we choose Wilcoxon

• a worked example

Answer 5

in this worked example we are looking for a difference in anger scored before and after an anger management programme

It is a REPEATED MEASURES DESIGN (related)

the data is ORDINAL as angers cores are based on a subjective “unsafe” self - report questionnaire

Question 6

what are the two different hypotheses

Answer 6

Alternative hypothesis: there is a difference in young offender’s scores on an anger questionnaire before and after treatment (non- directional two- tailed)

NULL HYPOTHESIS: there is no difference in young offender’s scores on anger questionnaire before ad after treatment

Question 7

what are the difference steps in Wilcoxon test

Answer 7

Step 1: calculate a difference and rank the difference

Step 2: working out the value of T

Step 3: The calculated and critical value

Question 8

how do we work out the calculated value

Answer 8

the calculated value of value of T is the sum of the less frequent sign

Question 9

how do we know if the results is significant

Answer 9

if the critical value of T is more than the calculated value, then the results are significant

this means we can accept the alternative hypothesis and reject the null hypothesis