EXERCISES AND PROJECTS

After each description of a process given below, indicate the level of measurement (nominal, ordinal, interval, or ratio) implied by the statement.

1. Asking people which television network they watch the most often: ABC, CBS, NBC, Fox, or CNN.__________
2. Asking people how much time during the week they spend watching television news programs: more than 10 hours, 7 to 10 hours, 3 to 6 hours. __________
3. Rating members of Congress as very liberal, slightly liberal, moderate, slightly conservative, or very conservative. __________
4. Asking members of Congress how much campaign money they have received from political action committees (PACs). __________
5. Asking members of Congress whether the amount of money they have received from PACs is more than $100,000, between $50,000 and $100,000, or less than $50,000. __________
6. Asking members of Congress the source of most of their PAC contributions: labor unions, corporate management, medical associations, or civil rights groups. __________
7. Asking members of Congress which of the following they feel is the most important problem facing the country today: the budget deficit, the threat of nuclear war, crime, poverty, or teen pregnancy. __________
8. Asking members of Congress if they think the amount of money the government is spending on AIDS research should be increased, kept the same, or decreased. __________
9. Checking the Congressional Record to see how many times in the past year members of Congress were absent from roll-call votes. __________
10. Asking members of Congress to indicate whether they are Democrats or Republicans. __________

You are in charge of a study of the attitudes of adult males toward affirmative action programs for women. One-half of the sampled males are interviewed by female interviewers, the other half by male interviewers. The results of your survey show that males interviewed by female interviewers have a higher approval of affirmative action programs than the subjects interviewed by male interviewers. Assuming that all subjects answered the same questions, what can you most likely conclude about your measure of affirmative action attitudes?

You are involved in a study of gender bias in state political party organizations. You decide to measure your key variable, gender bias, by (a) attending party organization meetings and observing whether conversations between women and men seem to be friendly or hostile, and whether men seem to pay attention to speeches made by women; and then (b) examining state party directories and calculating the percentage of leadership positions held by women. Which measurement method is more reliable? Which is more valid? Explain your answer below.

In considering levels of measurement, we often think that variables can be measured at only one level. However, it is frequently possible to measure a given variable at two or more levels. For example, although we usually think of age as a ratio variable, we could transform age into an ordinal variable by placing individuals into several age groupings, such as 18, 19-25, 26-40, and so on. Your task is to indicate whether (and if so, how) each of the following variables might be measured at different levels. In some instances, you may have to use your imagination!

1. Smoking

   Nominal:

   Ordinal:

   Ratio:

2. Gender

   Nominal:

   Ordinal:

   Ratio:

3. Socioeconomic status

   Nominal:

   Ordinal:

   Ratio:

4. Intelligence

   Nominal:

   Ordinal:

   Ratio:

5. Political modernization

   Nominal:

   Ordinal:

   Ratio: