Scope and Methods Assignment 4. Spring 2004.

Due in class on April 28.

Question 1.

There is a common hypothesis in political science that expects that income is positively related to the likelihood of voting in elections: as one's income increases, the likelihood of voting increases.

A quick survey after the last election revealed that 100 of 500 "low income" people voted, 500 of 1000 "middle income" people voted, and 400 of 500 "high income" people voted. Report these findings in a contingency table. What does the table show about the hypothesis?

We also should expect that education influences the likelihood of voting: as one's education increases, the likelihood of voting increases. Education should also influence income: as one's education increases, one's income increases. Thus we should include education as a control variable for our zero-order relationship above, between income and voting. (We'll measure "education" as either having a college education or not.) Given these expectations, what type of control variable is education? The table below shows he same people, now with the control of education added to the analysis. What does that table below show? HINT: Change these numbers to proportions.

	College			No College
	Low	Medium	High	Low	Medium	High
Voted	80	400	320	20	100	80
Did Not	10	13	3	390	487	97

Questions 2. Measures of Association

One hypothesis about politics is that last year men were more likely to support going to war with Iraq than women were. Using the (fictional) data presented in the table, calculate the appropriate measure of association to examine whether or not this hypothesis is correct.

	Male	Female
Supported War	375	250
Opposed War	125	300