1

CMNS260: Empirical Methods for Communication Research,

Exercises (Practice and Study Tips for the Final Exam)

The final examination will continue to have primarily close-ended questions (multiple choice, matching definitions, true-false and short answer questions. As well there will be

  • questions that require you to read tables and discuss the patterns you see in words;
  • causal diagrams;
  • and possibly, a short essay-style question.

The examination covers all material in the readings, lectures, labs and tutorials but will emphasize material that was either not well-done in previous quizzes or material that has not been treated in much depth in previous quizzes.

Suggested Strategy to Prepare for the Examination:Review the practice exercises and the results of the last two quizzes to determine areas you need to review in depth. Re-read the textbook and review the lecture notes. The glossary of terms (pp. 353-363) provide handy definitions you can use to test your knowledge of terminology.

Sample Questions

1. Here is data collected of a study of the age distributions of two groups of people waiting at different bus stops, one group in front of a bar and the other in front of an ice-cream parlour:

Ages of Group near bar: 25, 26, 27, 30, 33, 34, 35 (mean age=30 yrs)

Ages of Group near ice-cream parlour: 5, 10, 20, 30, 40, 50, 55, (mean age=30 yrs)

a. Name two other measures of central tendency and give them for each distribution.

b. Why are measures of central tendency inadequate for comparing these two distributions? Be specific.

c. Suggest another type of measure which could be used (be specific).

2. The following sets of numbers represent grades obtained by two different classes.

Class #1: 45 54 55 60 65 65 65 67 67 68 70 72 78 79 83 86 86 90 91

mean=70.7, median= 68, mode=65 standard deviation=13.0

Class #2: 40 50 53 54 60 65 66 66 66 69 70 71 74 78 82 84 86 90 95

mean=69.5, median=69 mode= 66 standard deviation=13.2

Which class has marks that are more spread out? Explain briefly how you know this specifying which measure you used to figure it out.

3.Suppose I find an inverse relationship between parental income and juvenile delinquency. Which of the following statements could apply?

a.The richer the parents the more likely they are to have children who are juvenile delinquents.

b.Juvenile delinquents are more likely to have poor parents.

c.Juvenile delinquency increases with parental income.

d.As parental income decreases, juvenile delinquency decreases.

4.Referring to the above question, when I control on whether the parents have an alcohol problem or not the relationship disappears. In this case which of the following statements could apply? (You may choose more than one.)

a.The percentage of juvenile delinquents from rich families is close to the percentage from poor families when neither of the parents are alcoholics.

b.The percentage of juvenile delinquents from rich families is close to the percentage from poor families when one or more parent is an alcoholic.

c.A much larger percentage of delinquents with alcoholic parents come from rich families than from poor families with alcoholic parents.

d.A much smaller percentage of delinquents with non-alcoholic parents come from rich families than from poor families with non-alcoholic parents.

5.In the previous case a relationship between two variables disappears with the introduction of a third control variable. Identify and discuss which of the patterns described in the elaboration paradigm could this be? What assumptions would you need to make?

Topic: Reading Tables Presenting Standardization & Rates

Example 1: : (See Neuman & Robson, p. 122 for a discussion of this example.) Use the table below to answer the questions below the table:

a) Which statement(s) describe what the numbers presented in first column on the left represent?

  1. The number of gold, silver and bronze medals won by athletes from each country.
  2. The number of gold medals won by athletes from each country.
  3. The ranking of countries determined by the number of gold medals won by athletes from each country with the countries listed according to their performance in the groups with ties.
  4. The ranking of countries determined by the number of gold medals won by athletes from each country with the countries listed with no explanation of the order of their listing in the groups with ties.
  5. The countries ranked in the top six according to the proportion of gold medalists taking into account the size of the population

b) Briefly describe in a few sentences how the ranking presented in the 4th column from the left was done.

c) What the strengths and weaknesses of the reasoning (logic) that led researchers to find a second method for ranking countries.

d) What is the best technical term used to describe this type of measure that takes population size into account in the rankings?

  1. A scale
  2. A dependent variable
  3. A standardized rate
  4. A sampling element
  5. A ratio measure

Example 2:

A television news show featured a story about eating disorders of female university students. The story was based on findings from a government-funded health survey. For the purposes of this discussion let us assume that the sample was representative and the measures valid.

According to the report, 400 students at North University suffered from eating disorders. In addition there were 200 cases at East U., 180 cases at West U. but only one case at Royal University (which is a military school known for strict discipline). Reporters concluded that North had the worst problem because it had the most cases.

North U. had spent a great deal of money on a new women’s center aimed at helping female students with problems like this. Critics claimed that these new statistics showed that the woman’s center had just wasted the money. One person also suggested that all the universities be made to follow the strict discipline at Royal U. since only one student attending that school had eating disorders.

Imagine that you are called in to analyze the situation and you obtain the statistics about each university presented in Table A below. Use them in answering questions a-e.

Table A: Enrollment Information and Eating Disorders Amongst Female Students in Universities

University / # of female students with eating disorders / total enrollment
(# of students) / # of female students / # of male students / crude rate of eating disorders (# of female students with disorders divided by total # of students)* / rate of eating disorders among female students
(# of students with disorders divided by # of females)*
North U. / 400 / 10,000 / 6,000 / 4,000 / .04 / .07
East U. / 200 / 4,000 / 1,600 / 2,400 / .05 / .13
West U. / 180 / 3,000 / 1,200 / 1,800 / .06 / .15
Royal U. / 1 / 500 / 5 / 495 / (.002) / .20

* these have been rounded to nearest .00th in most cases.

a)Fill in the following table by writing the names of the universities in the empty squares.

Table B: Ranking of Universities According to Statistics on Statistics on Eating Disorders of Female Students

Ranking
(1=most, 4=least) / Universities Ranked by Total Number of Cases / Universities Ranked by Crude rate of eating disorder (for total student population at each school) / Universities Ranked by Rate of eating disorders among female students
1 (highest)
2
3
4 (lowest)

b) What is wrong with the idea of using Royal University as a model for reducing eating disorders in the female students at other universities?

c) What is a possible problem in including Royal U. in the rankings with the other universities?

d) The director of the women’s center at North U. wants to make a statement to the press explaining that even though North’s student population has the most cases of eating disorders the survey research findings show that the situation at other universities is worse. Explain how information in tables A and B could be used by the director in an analysis of the situation. Try to make at least two points using specific information from the tables in your discussion.

e) Suppose that you wanted to compare the situation at North U. with the situation at East U. by calculating the number of cases North U. would have if it had the same rate of eating disorders per female student at East U. How exactly would you do the calculation? (Either do the calculation or give the numbers you would use to calculate this indicating what operations you would perform to do it.)

Topic : Indices (plural form of Index) and Composite Measures

Maclean’s Magazine has been conducting an annual study of Canadian universities for years. Until recently, part of the study focuses on performance measures, which the magazine’s research team uses to rank different types of universities. The magazine used two main sources of information for the rankings :

  1. responses to a questionnaire filled out by administrative staff (not faculty members or students) of participating universities and
  2. responses to a survey about the reputation of universities sent to high-school guidance counselors and executives of corporations.

The performance measures used in ranking “primarily undergraduate” universities in the 1990s are presented in the table below. Use this table, your readings and your knowledge of SFU to answer the questions below.

GROUPING of MEASURES / % OF FINAL SCORE
STUDENT BODY / 21
-average grades of incoming students / 12
-proportion of incoming students with 75% or more / 3
-proportion of incoming students from out of province / 1
-graduation rates
(% full-time undergrads in second year who graduate) / 2
-national academic awards won by students / 3
CLASSES / 18
-class sizes at 1st and 2nd year levels
(1-25, 26-50, 51-100, 101-250, 251-500, 501 plus) / 7.5
-class sizes at 3rd and 4th year levels / 7.5
-% 1st year courses taught by tenured or tenure-track profs / 3
FACULTY / 17
-% with Ph.D’s / 3
-national academic awards won by professors / 3
-success of eligible faculty in winning federal grants
(arts and social sciences) / 5.5
-success of eligible faculty in winning federal grants
(medicine and sciences) / 5.5
FINANCES / 12
-available money for current expenses per student / 3.33
-% of budget spent on student services / 4.33
-% of budget spent on scholarships and bursaries / 4.33
LIBRARY / 12
-number of volumes per total number of students / 4
-% of university operating budget spent on library services / 4
-% of library budget spent on updating collection / 4
REPUTATION / 20
-number of gifts to university from alumni in past 5 years / 5
-poll of guidance high school guidance counselors from 2648 schools and C.E.O’s (chief executive officers) of corporations / 15
TOTAL / 100

a. What percentage of the score for reputation is based on the opinions of high-school guidance counselors and executives of corporations? ______

b. What percentage of the measures of excellence is based on the opinions, attitudes and testimony of university students attending the universities under study?

c. Although Maclean’s Magazine claims that the measures reflect differences in quality of the universities studied, many of the measures focus on quantitative differences that may not fully reflect the theoretical concept of quality in higher education. Choose two examples of measures used in this study which may not be valid indicators of academic excellence. Explain why, indicating what kind of validity might appear to be lacking for each example.

Look through your textbook and lecture outlines for information about the different types of validity. Then critically assess the measures used, relating them to specific types of validity.

Topic : Practice Reading, Describing and Analyzing Cross-tabulations

You may also be asked to comment on tables, describing bivariate and trivariate crosstabs and analyzing patterns as you were asked to practice in Assignment 6.

Topic: Critique of Survey Questionnaire Design

Pre-tests are a useful way of locating problems and shortfalls in data collection tool. The class pre-tested a survey on The Ethics and Etiquette of Communication Media use and used the findings to analyze data about attitudes and practices. On the basis of your work with the survey results prepare a short description of problems with the questionnaire and ways of improving the questionnaire that addresses the following points:

  • coverage of topics
  • question order
  • a specific suggestion for improvements to a specific existing close-ended question or a proposal for a new question. (Include
  • the conceptual and operational definitions of the variable that the question is designed to measure,
  • the precise wording of question,
  • the response categories and
  • plans for use of the data gathered.