Exercises 31310.11 An investigator is interested in the relationship between alcoholism and a childhood his-
tory of attention deficit disorder (ADD). He has collected the following data, where a 1
represents the presence of the relevant problem.
ADD: 0100110001001001
Alcoholism: 0100010001100001
ADD: 1100000001001000
Alcoholism: 0100000001001010
a. What is the correlation between these two variables?
b. Is the relationship significant?10.12 An investigator wants to arrange the 15 items on her scale of language impairment on the
basis of the order in which language skills appear in development. Not being entirely
confident that she has selected the correct ordering of skills, she asks another professional
to rank the items from 1 to 15 in terms of the order in which he thinks they should appear.
The data are given below.
Investigator: 12345678 9101112131415
Consultant: 1324756810 91112151314
a. Use Pearson’s formula (r) to calculate Spearman’s.
b. Discuss what the results tell you about the ordering process.
10.13. For the data in Exercise 10.12,
a. Compute Kendall’s.
b. Test for significance.10.14 In a study of diagnostic processes, entering clinical graduate students are shown a
20-minute videotape of children’s behavior and asked to rank order 10 behavioral events
on the tape in the order of the importance each has for a behavioral assessment (1 5 most
important). The data are then averaged to produce an average rank ordering for the entire
class. The same thing was then done using experienced clinicians. The data follow.
Events: 123456 78910
Experienced clinicians: 132754 86910
New students: 2416531087 9
Use Spearman’s to measure the agreement between experienced and novice clinicians.
10.15 Rerun the analysis on Exercise 10.14 using Kendall’st.
10.16 Assume in Exercise 10.14 that there were five entering clinical students. They produced
the following data:
Student 1: 142653 9107 8
Student 2: 4325711086 9
Student 3: 152643 8107 9
Student 4: 2517431086 9
Student 5: 251463 97810
Calculate Kendall’s Wand for these data as a measure of agreement. Interpret your results.rsrsttrsExercises 313