Hypothesis Testing: Consistency and Confusion---Factors Related to Grade Performance

• Activity Number: 38
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Education
• Author(s): John Barroso*+
• Companies: University of Pittsburgh
• Keywords: grade performance ; logical statistical reasoning ; GPA ; structural similarities theory ; psychological variables ; hypothesis testing

Abstract

This research study assesses the role of relatively "unappreciable variables" (mostly self-reported perceptions such as number of friends, household mathematical knowledge, dislike of statistics, in love in a romantic relationship, self-definition of beauty, self-definition of rational decision making, self-definition of wisdom, drive towards higher pay, sexual orientation and drug use) in producing a significant association with grade performance. The study assesses consistencies and inconsistencies in statistical learning. A consistency exists when related questions receive consistently correct or consistently incorrect answers. An inconsistency happens when two related questions have contradicting answers. The sample consisted of eighty-eight statistics students who answered a 22-question questionnaire and took a mock quiz very similar to a real quiz. The statistical techniques of Regression and two-by-two Chi-Square tables were used. Among the findings are four significant regression models in predicting grade performance, three unappreciated variables significantly associated with grade performance, and six Chi-Square tests showing which pairs of concepts receive consistent or inconsistent statistical logical reasoning. The paper's theoretical framework is based on Structural Similarities Theory and the findings are of relevance to understanding not why but where students deploy consistent or inconsistent logical statistical reasoning to solve related concepts when taking essay quizzes in statistics.