Exam 5: Making Controlled Comparisons

Rival explanations undermine researchers' ability to evaluate the effect of the independent variable on the dependent variable.

Matching methods attempt to replicate random assignment in an observational setting.

Cross-tabulation analysis may be used only when both variables are measured at the ordinal level.

A researcher hypothesizes that individuals differ in their support for free speech based upon partisanship. He controls for gender and finds that women are more supportive than men for both Democrats and Republicans. What's going on here?

In a study of partisanship and attitudes toward social welfare spending while controlling for gender a researcher finds that women are 15 points more likely than men to support social welfare spending. This is an example of ______.

A method of measuring respondents' attitudes from 0 to 100 is called a ______.

In political research, controlled effects are summarized by a(n) ______ relationship.

When the partial effect must be described separately for each category of a control variable what type of relationship exists?

The rule of direction for nominal relationships helps us identify ______.

We are able to summarize a partial effect using one number when the relationship is ______.

A(n) ______ summarizes a relationship between two variables after accounting for a rival variable.

An outlier is a pretreatment variable that is related to both the treatment and the outcome.

A zero-order relationship summarized an overall relationship between variables.

A relationship between a causal variable and a dependent variable within one value of another causal variable is known as a ______.

Determining whether an independent variable is positively or negatively related to a dependent variable is done using the partial relationship or partial effect.

An overall association between two variables that does not take into consideration other possible differences between the cases being studied is known as a zero-order relationship.

If age and gender together help to explain attitudes regarding welfare policy, it indicates that a(n) ______ is present.

All additive relationships are straightforward and have symmetrical quality.

When you're analyzing how values of an interval-level variable vary among groups, you use a mean comparison table.

If there is a clear rival explanation, a controlled comparison design does not work well.