Which non-orthogonal contrast compares the mean of each condition to the mean of all preceding conditions combined?

Which non-orthogonal contrast compares the mean of each condition to the mean of all preceding conditions combined?

• A. Deviation contrast
• B. Difference contrast
• C. Direct oblimin
• D. Discriminant analysis

Answer: (B)

The question is about how a specific contrast tests hypotheses across ordered condition means. A difference contrast is designed to test whether each successive condition differs from the average of all preceding conditions. In other words, for a sequence of ordered conditions, you compare the mean of the next condition to the mean of all previous ones. For example, with five ordered conditions, the first contrast compares the second condition to the first; the second compares the third to the average of the first two; the third compares the fourth to the average of the first three; and the fourth compares the fifth to the average of the first four. This pattern captures incremental change as you move along the ordering. These contrasts are non-orthogonal because the later comparisons reuse information from earlier levels, so their sums of squares aren’t independent. The other options don’t describe this specific pattern: one contrasts each level with the grand mean, another is a rotation method used in factor analysis, and another is a classification technique, not a contrast for comparing means.

The question is about how a specific contrast tests hypotheses across ordered condition means. A difference contrast is designed to test whether each successive condition differs from the average of all preceding conditions. In other words, for a sequence of ordered conditions, you compare the mean of the next condition to the mean of all previous ones.

For example, with five ordered conditions, the first contrast compares the second condition to the first; the second compares the third to the average of the first two; the third compares the fourth to the average of the first three; and the fourth compares the fifth to the average of the first four. This pattern captures incremental change as you move along the ordering.

These contrasts are non-orthogonal because the later comparisons reuse information from earlier levels, so their sums of squares aren’t independent. The other options don’t describe this specific pattern: one contrasts each level with the grand mean, another is a rotation method used in factor analysis, and another is a classification technique, not a contrast for comparing means.