Two batteries of statistical tests were performed to compare the Moogie Math Scores with actual state test scores. The first battery used the test of simple regression, while the second battery employed the method of weighted least squares. These tests focused on three variables for each test case: state test score, Moogie Score, and Moogie Confidence Level. However, before I continue, some definitions are in order.

When using the Moogie Math software, a Score is actually retained for each question in the Moogie Math database that is arbitrarily set to 18. Each time the student correctly answers the question, the Score for that question is increased by one. When the question is answered incorrectly, the Score is decreased by two. The Scores for all the questions in any given objective (or learning outcome) are averaged to create a Score for that objective. Similarly, the Scores for all the questions in a single strand (or topic, such as Algebra) are averaged together to yield a Score for that strand. Finally, the Scores for the five strands are then averaged to produce an Overall Score.

The Moogie Confidence Level, which is also reported by the software, measures the percentage of questions in a strand that a student attempts to answer. For example, if there are 100 questions in the strand and a student tries 75 of them, then the student's Confidence Level for this strand would be 75%. If, however, the student only attempts one question, even if the question is re-tried 50 times, then the student's Confidence Level is only 1%. It is only logical that faith should not be placed in a student's Score if a low percentage of questions is attempted.

In each battery of tests, students' Scores for the five strands as well as their Overall Scores in the Moogie Math database were compared to their actual state test scores. The data for this study were obtained from two high school classes and were combined into one single dataset. The purpose of the tests was to reject the null hypothesis at the .05 level of significance chosen apriori. The null hypothesis is that there is no correlation between Moogie Scores and actual state test scores.