Exam 17: Testing for Differences Between Two Groups or Among More Than

SPSS will report analysis of differences tests in terms of t tests.

ANOVA is called a "signal flag" procedure, meaning that if it shows significance:

Research determines a significant difference between cold remedy preferences depending on symptoms to be relieved. When the marketing manager markets several different remedies, each designed to reduce a different symptom, we can say the differences found in the research are "actionable."

If we are examining the SPSS output for testing the differences between two means from independent samples (Independent Samples t Test)and we show a Sig. of .001 for Equal variances assumed under Levene's Test for Equality of Variances, this means we:

When testing the difference between two percentages from two independent samples, a z value larger than 1.96 indicates the difference is NOT significantly different.

Analysis of variance is a "signal flag" procedure. By this, we mean that it alerts the researcher to the fact that there are many differences that are statistically significant and it flags him or her to those differences which are most important, those which should be acted upon first.

If the F value in ANOVA produces a significantly high p value (aka "Sig." in SPSS)of 10 or more, then it is appropriate to proceed with an ad hoc test.

Different tests are required to test the differences between the means of two samples depending upon whether or not the two samples are independent or are paired.

The t value is used for many tests instead of the z value because:

Which of the following is NOT true of a differences test?

When a researcher is determining if two groups are statistically significant, he or she is considering the two groups as two separate populations. The question is whether or not the two different populations':

P values are identified by what terms on computer output by statisticians?

A group comparison table is often used to summarize significant differences when reporting group differences to clients.

Assume you have two questions of the same format, a 5-point scale measuring importance level. One question measures importance of prices in the selection of a restaurant and the other measures food quality. If you want to know if the two mean scores to each question are significantly different, you would use:

The t test is the statistical inference test to be used with small sample sizes less than or equal to 30.

Joy Ward is the director of marketing at Helmsley College. She has been studying marketing research data collected on a national sample of high school seniors who are planning on attending college. Joy is trying to determine what appeal she should use in a direct mail campaign that will be targeted at students with high SAT/ACT scores who live within 500 miles of Helmsley. She is intrigued with the marketing research data that measures the students' ratings of importance on a number of factors affecting their decision to choose a particular college. Some of the factors are (1)programs highly valued in the job market, (2)small campus atmosphere where professors know students' names, (3)ample opportunities for an active campus life (i.e., football, campus entertainment for students, etc.), (4)individualized programs designed around students' needs and interests, and so on. There are eight different factors and each was rated on the same 5-point importance scale ranging from "Very Important" to "Very Unimportant." Joy knows that Helmsley could ethically use either the "small campus" appeal or the "individualized program" appeal. The small campus appeal has a mean score of 4.3 on the importance scale, while the individualized program appeal has a mean score of 3.95. Because both of these numbers indicate that these are important factors for students, she wants to know if the small campus appeal is really more important than the individualized appeal in the total college bound population. Which of the following tests should Joy run?

The Duncan's Multiple Range test is a post hoc test that allows the researcher to determine among which pairs of means significant differences exist.

Which of the following commands would be used to test for the significant differences between two percentages?

If we wanted to test for the mean likelihood that shoppers will shop again in different departments such as home and garden versus sporting goods, versus automotive, versus electronics, we would use:

Unfortunately, SPSS cannot run the paired samples t test.