Question 2 (16 marks) After the researcher completed the analysis of the data given in Question 1, the lecturer told her that the marking of those 24 scripts (completed assignments) was actually conducted in the following way. The 24 scripts were sorted into four bundles first, one bundle for each passage (ie, having six scripts in each passage bundle). The lecturer then randomly selected one script from each of these four passage bundles to form six groups of four, ie, each group with one and only one script from each passage. Each of the six groups was then given by random to one of the six markers to mark. The data after including marker's information are given in the table below. (Note: May use R for any port below when possible.) Passage Marker A B c D lrisa 22 21 27.5 16.5 Ben 32 22.5 20 18.5 Andrew 29 17.5 23 19 Sue 25 25 20 20 Kate 31 23 33 25 Sam 32.5 26 32 21.5 a) What design has been used here (or, can be considered as)? b) Write out the statistical model for this experiment, and associated constraints, if any. c) Test at 5% significance level if there is any difference in average mark across the four passages? (Note: Need to carry out an appropriate ANOVA first.) d) If one wishes to use Bonferroni's method to investigate simultaneously the three contrasts formed in part f of Question 1 at 5% significance level, what will be the significance level for each of these three contrast comparisons? For these three comparisons, you are ONLY required to carry out and show an appropriate hypothesis test for the first contrast as one of the three simultaneous comparisons, to investigate whether it is easier to translate passage A or C than translating passage B or D to English (ie, N0 need to do or show the other two tests here). Interpret the result