A researcher interested in the effect of status on compliance confronts people on the street with a request to answer a "short" set of questions. Her survey actually consists of a lengthy series of items of increasing sensitivity, which she continues to ask until the respondent refuses to go on. The number of questions the respondent answers provides a measure of compliance. As a measure of status, one-third of the respondents are told that the researcher is from a prestigious university, one-third are told that she is from a non-prestigious university, and the final third are told nothing about her university affiliation.

Which statistical procedure would you apply to test for differences in extent of compliance by status of university affiliation?

		Chi-square
		t-ratio
		ANOVA
		Median test