This is a worksheet for RESEARCH METHODS IN POLITICAL SCIENCES using microcase.

Chapter Ten Worksheet

**Exercise 1:**

Briefly describe the pattern in the results displayed in this table.

____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Because both of these variables are ordinal, we need ordinal statistics. Write the significance level.

Prob. = ___________

Is this relationship statistically significant? Yes or No. _____________

Do we reject the null hypothesis? Yes or No. _______________

Write the value of Kendalls tau-b. Kendalls tau-b = ___________________

**Exercise 2:**

Is there a pattern in those results? Do African Americans oppose capital punishment more than do white Americans? Yes or No. ___________________

Because race is a nominal variable, we need to use nominal statistics. Based on the chi-square test of statistical significance, what is the significance level here? Prob. = __________________

Is this relationship statistically significant? Yes or No. _________________

The null hypothesis is that there is no relationship between support for capital punishment and race? Do we reject the null hypothesis? Yes or No. _______________

Next, write in the value of lambda and the value of Cramers V for this relationship.

Lambda = _________________ Cramers V = _____________

**Exercise 3:**

Compare the column percentages horizontally. Is there a pattern? That is, do males and females differ substantially in their responses to this question about allowing an abortion? Yes or No. ____________

Because gender is a nominal variable, we need to use nominal statistics. Based on the chi-square test of statistical significance, what is the significance level here? Prob. = ____________________

Is this relationship statistically significant? Yes or No. ________________

The null hypothesis is: There is no relationship between support for allowing an abortion and gender. Do we reject the null hypothesis? Yes or No. _____________________

**Exercise 4:**

Is it more appropriate here to use nominal or ordinal statistics? Circle one.

Nominal Statistics Ordinal Statistics

Present the significance level for this relationship. Prob. = _________________

Is this relationship statistically significant? Yes or No. _______________

Do we reject the bull hypothesis? Yes or No. __________________

Which measure of association would be most appropriate here? Circle one.

a. Lambda

b. Cramers V

c. Kendalls tau-b

d. Kendalls tau-c

Present the value of the measure of association you selected above. ___________________

Is this relationship positive or negative? Circle one.

Positive Negative

Are the results in accord with the hypothesis? Yes or No. ___________________

Give a P-R-E interpretation of these results.

____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

**Exercise 5:**

In this situation, is there a relationship between political distrust and the belief that elections cause government to pay attention to the people? Yes or No. _______________________

If you answered yes above, briefly describe the kind of relationship that exists between political distrust and the belief that elections cause government to pay attention to the people.

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Both variables are ordinal, so we will use ordinal statistics. The significance level for ordinal variables is presented on the last line of the statistics. Write in the significance level. Prob. = _______________

Is this relationship statistically significant? Yes or No. ________________

If we say that there is a statistically significant relationship between trust in government and the belief that elections cause government to pay attention to the people, what would that mean? That is, what decision are we making concerning these results based on a sample?

____________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

The null hypothesis in this situation is that there is no relationship between trust in government and the belief that elections cause government to pay attention to the people. Based on these, results, should we reject the null hypothesis? Yes or No. _____________________

Now lets look at the strength and direction of the relationship. Because both variables are ordinal, we will use Kendalls tau. Specifically, because both variables do not have the same number of categories, we will use Kendalls tau-c. Write in the value of this measure of association.

Kendalls tau-c = _________________

Is this relationship positive or negative?

Positive Negative

Thus people who feel confident that elections make the government pay attention to them are more likely to have. Circle one.

Higher political distrust Lower political distrust

Give a P-R-E interpretation of Kendalls tau in this situation.

**Exercise 6:**

Do the results mostly support the hypothesized pattern? Yes or No. ____________________

For which of the following groups are the results least in accord with the hypothesized pattern? Circle one.

Democrats Independents Republicans

Write in the significance level in the space provided. Prob. = ______________________

Is this relationship between part identification and 2008 presidential voting preference statistically significant? Yes or No. ______________________

The null hypothesis here is that there is no relationship between political identification and 2008 presidential voting preference. Do we reject the null hypothesis? Yes or No. ___________________

Write the value of lambda. Lambda = ____________________

According to lambda, by what percentage have we reduced the number of errors in predicting the dependent variable by knowing the independent variable? __________%

On the basis of those results, what would you conclude about the extent to which political party identification is related to voting preferences in the 2008 presidential election.

_______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

<