Introduction to Research Design and Statistics
Pseudo-Assignment #4
Part I:
1. Generate a contingency table for Female by Race.
a. On the print-out, circle the marginal distribution for Female.
b. On the print-out, circle the joint distributions for Race and Female.
c. Which gender has the largest percent of Race = 1? Circle on print-out.
d. How many Race = 1 students are female?
e. State the null and alternative hypotheses for the chi-square test for this table.
f. What is the observed value of chi-square and what are the degrees of freedom
(circle them on the print-out)?
g. What is the critical value of chi-square?
h. State your conclusion.
i. Interpret the results.
2. Generate a contingency table for Prog by Schtyp.
a. What are your observations
regarding the percent of vacational at each type of school?
b. Include the value of the percents in your response.
c. The observed value of chi-square.
d. The critical value of chi-square.
e. The degrees of freedom.
f. State your conclusion.
g. Interpret the results.
3. One item on a survey given to a random sample of 1,100 high school students asked, "How often do you use
marijuana?" The choices were a) Never used it,
b) Tried it once, c) Use it infrequently, d) Use it often.
Below are the responses to the item:
Marijuana use | Frequency
|
---|
Never used it | 380 |
Tried it once | 210 |
Use it infrequently | 330 |
Use it often | 180 |
A similar survey was done in 1972 found the following results: 21% selected never tried it, 12%
selected tried it once, 43% selected use it infrequently, 24% selected use it often.
Test the pattern of frequencies between the current survey and the one given in 1972. Include:
a. The research hypothesis.
b. The null and alternative statistical hypotheses.
c. The observed value of chi-square.
d. The critical value of chi-square.
e. The degrees of freedom.
f. State your conclusion.
g. Interpret the results.
Part II:
In our next analyses, you will be focusing on the relationship between READ and MATH for each
gender. In other words, is there a relationship between performance on the reading test and
performance on the math test. How much of the variance in performance on the reading
test appears to account for performance on the math test for each gender? Circle
each of the results on your printouts.
1. Based on the scatter plots, how would you describe to association between READ and MATH
for each gender?
2. What is the correlation between READ and MATH for each gender.
3. What is the value of the coefficient of determination for each gender?
4. Given what you know about the characteristics of male and female students, what are
some of your thoughts regarding the difference between the two correlation coefficients?
5. The Spearman Rho Coefficients for MATH with SES at for males is .2417 and for females is
.3334. What are your thoughts regarding why this difference in Spearman correlations
occurred?
Intro Home Page
Phil Ender, 30Jun98