6. (10 points) MSU researchers have employed five questionable indicators of the importance of pollution problems. The codes of these five indicators are: 1 for Very Serious Problem, 2 for Somewhat Serious Problem, 3 for Small Problem, and 4 for No Problem. These five questionable indicators are listed in the five rows in the table below. The researchers believe that these five indicators are measured at the ordinal level of measurement, so that is the key validity question.

A construct-criterion validity test was performed on each of these five questionable indicators. The well established indicator was preference for state government spending on environmental programs, whose codes ranged from a low of 1 for spending less than current expenditures to 2 for spending the same amount as currently to 3 for spending more than the current budget. That well established indicator is clearly measured at the ordinal level . Each cell entry in the table below indicates how much each of the four groups at the top wanted to spend on environmental programs.

For example, in the first row, the 2.1 value in the second column indicates that among those Mississippians who rated drinking water pollution as a Small Problem, their average preference was that government should spend about the same as it currently was on environmental programs. The 2.9 value in the last column of that same row indicated that among those Mississippians who rated drinking water pollution as a Very Serious Problem, their average preference was that government should spend more than it currently was on environmental programs. Our construct validity test expects to find that as people view an environmental problem as more serious, they will desire to spend more money on environmental programs.

Do any validity problems exist with any of these five questionable pollution problem indicators? If so, indicate which pollution problem item or items had validity problems, and indicate what categories of that item or items demonstrated validity problems. Also, circle whatever adjacent categories of an item demonstrated validity problems.

Pollution Problem Item (questionable items)	No Problem	Small Problem	Somewhat Serious Problem	Very Serious Problem
Drinking Water Pollution	2.4	2.1	2.5	2.9
River/Lake Pollution	2.5	2.2	2.6	2.7
Air Pollution Problem	2.3	2.2	2.6	2.5
Litter Problem	2.2	2.3	2.4	2.5
Solid Waste Disposal Prob.	2.3	2.2	2.6	2.5

Note: Cell entries are the means of the Environmental Programs state spending item, a well established indicator whose codes ranged from a low of 1 for spending less to a 3 for spending more. The five questionable indicators are listed at the left, and each indicator has a separate row.

8. (10 points) Political researchers are interested in how political party activists think about political issues. They conduct a regional public opinion poll asking Democratic and Republican activists their attitudes toward six issues: abortion; school prayer; death penalty; education spending; health care spending; and social welfare spending. They theorize that public attitudes on policy issues constitute a single dimension of liberalism versus conservatism--that a citizen consistently either supports or opposes government programs that held the socially disadvantaged or unpopular individuals, such as the poor, women, atheists, and criminals. They therefore creates a summary state policy "ideology" scale by combining each person's score on each of these six indicators, believing that all six questions measure the same concept or dimension. All indicators are measured in a liberal to conservative direction.

One political science researcher believes that there may be a validity problem with a single scale of state policy ideology. She argues that there may be different components to public policies, and that some citizens may support certain types of government programs but not others. Conducting a Convergent-Discriminant Validity Test, she generates a correlation matrix to study how peoples' responses on the six indicators are interrelated.

CORRELATION MATRIX	Abortion	School Prayer	Death Penalty	Education Spending	Health Care Spending	Social Welfare Spending
Abortion	1.0
School Prayer	.75	1.0
Death Penalty	.9	.85	1.0
Education Spending	.14	.03	.15	1.0
Health Care Spending	.05	.04	.11	.92	1.0
Social Welfare Spending	.02	.04	.03	.86	.91	1.0

Is there any validity problem with forming a single scale that combines all six indicators of public policy into a single ideology scale, as the first researcher did? If so, what is the problem? Explain and discuss your answer, and refer to specifics in the correlation matrix. In your answer discuss how many dimensions of attitudes toward public policies there are, how you can determine that, and what they appear to pertain to.