Contrast and Pairwise Testing in MarketSight

Contrast Testing

Contrast testing compares each cell of a row variable to the group of other cells within the same row. MarketSight automatically selects the appropriate contrast test for you and displays significant results using yellow shading. For more information on how to interpret significance shading, see: Interpreting Significance.

If crosstab display options are set to display means for a row variable, MarketSight will compare each column’s mean to the aggregate mean of all other cells in the same row. Contrast tests for each cell are only run if higher level statistical testing indicates that a significant relationship exists between the column variable(s) and row variable. For example, MarketSight will only run a contrast test on a row variable displaying means if an F-test first indicates there is a difference between the different column means of the row variable.

If crosstab display options are set to display counts and/or frequencies (Column %) for a row variable, MarketSight will compare each cell’s frequency to the aggregate frequency of all other cells in the same row. Contrast tests will be run whether or not higher level statistical testing indicates that a significant relationship exists between the column variable(s) and row variable.

Pairwise Testing

Pairwise testing compares each cell to every other individual cell within the same row. MarketSight automatically selects and runs the appropriate pairwise test and displays significant results using blue shading. For more information on how to interpret significance shading, see: Interpreting Significance.

If crosstab display options are set to display means for a row variable, MarketSight will compare each column’s mean to every other column’s mean in the same row.Pairwise tests for each cell are only run if higher level statistical testing indicates that a significant relationship exists between the column variable(s) and row variable.

If crosstab display options are set to display counts and/or frequencies (Column %) for a row variable, MarketSight will compare each cell’s frequency to every other cell’s frequency in the same row. Pairwise tests will be run whether or not higher level statistical testing indicates that a significant relationship exists between the column variable(s) and row variable.

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.

Still need help? Contact Us Contact Us