You are a football coach interested in analysing play performance in three

different teams over a season. You chose the 10 top midfield or forward scorers

in each team and record: miles run per game, goals scored, playing position

(midfield, forward) and rank in the team relative to goals scored (top half,

bottom half). Raw data is given in Table A. Analysis of data is shown in the SPSS

output provided in the supplementary booklet.

a) You want to assess the association between “playing position” and “rank in the

team”, and decide to use a test applied to a contingency table, as shown in the

output. Which test is most appropriate? Justify your choice. What are the

statistical hypotheses in this test? What’s the p-value of the test? Would you

reject or fail to reject the null hypothesis? What statistical conclusion can you

draw from the output in this case?

b) You want to test if there is a significant difference between goals scored by

teams 1, 2 and 3. Outline why the Kruskal-Wallis test is appropriate in this

instance. Additionally state the statistical hypotheses under investigation and

with reference to the computer output fully outline what can be concluded from

the Kruskal-Wallis analysis.

c) You want to test if the means of “miles run per game” differ between teams and

use an ANOVA. What’s the factor being analysed and how many levels does it

have? What are the statistical hypotheses in this test? What’s the p-value of thetest? Would you reject or fail to reject the null hypothesis? What statistical

conclusion can you draw from the output in this case?

d) What complementary tests can be used after that used on the previous question?

Which one was used in the analysis shown in the output? Which means are

significantly different from which? What’s the level of significance of the

differences found?

e)

You want to study the correlation between the variables: “miles run per game”

and “goals scored”.

What correlation test should be used? Justify your choice.

What’s the correlation value? Is this significant? if yes, what’s the level of

significance?

f) Relative to the correlation studied in the previous question, what graphic would

you chose to display this data?

Sketch a graph of the correlation, showing the variables in the axis and the

tendency line. Use your own words to conclude about this correlation.