A Multifaceted Approach to Examine and Improve Undergraduate Introductory Statistics Education

• In Event: Innovation in Introductory Statistics Instruction: Student Attitudes, Motivations, and Peer Support as Key to Learning

Abstract

Objectives
Introductory statistics is often a required course for an undergraduate psychology degree, as well as being a pre-requisite for many graduate programs. Statistics courses typically cover descriptive and inferential approaches to data analysis. These courses rely on prior knowledge of basic mathematical skills, which is common roadblock to the successful completion of a statistics course for some students. Here, we present our work on the development of a measure to assess basic mathematical skills and examine the influence of demographic (gender, year in school, race/ethnicity, transfer status), psychological/behavioral (self-efficacy/outcome expectancy, self-reported help-seeking, and self-reported procrastination), and academic (basic mathematics skills) variables within a model that aims to explain the variance in undergraduate statistics course performance. We also introduce the use of computer-based simulations to explain abstract statistical concepts, while simultaneously providing students with a generalizable skillset that enhances their professional portfolio.

Perspectives
College-level statistics courses cover a range of topics, often with little time devoted to the details necessary to promote deep learning and critical thinking. Research has also shown that accessibility to quality education, anxiety related to help-seeking, disparities in self- efficacy and outcome expectancy among students, and demographic factors have a strong role in determining course outcomes. Therefore, a multifaceted examination of these various factors is necessary to understand differences in student performance and develop evidence-based practices to improve statistics education. Moreover, with psychology (and related disciplines) becoming more computationally oriented, our goal is to develop a comprehensive curriculum to help novice students learn statistics through intuitive computer coding and simulations.

Methods
We recruited undergraduate students enrolled in the required introductory statistics course for the psychology major from a four-year college in the Northeastern U.S. In our first study (Rabin et al., 2018), 414 students aided in the development and validation of a new tool (Math Assessment for College Students or MACS) that measures basic mathematics deemed critical for statistics competency. In our second study (Rabin et al., 2021) with 460 students, we used discriminant correspondence analysis to examine multiple variables in one multivariate model and quantify their contributions to differences in course examination performance.

Results
We successfully developed the 30-item MACS, which included computational questions (e.g., solving for an unknown variable in an algebraic expressions) and conceptual items (e.g., related to graphical interpretation and correct symbol notation). The MACS showed strong psychometric properties including factor structure and convergent validity. Subsequently, we found that higher examination grades were associated with higher scores on the MACS over and above all other psychological/behavioral and demographic variables included in the model.

Scientific Significance
Our research broadly examines multiple factors associated with academic success in an undergraduate introductory statistics course. This is an ongoing effort to promote statistics competency where our next step is to deploy and assess a free and open- source, novice-friendly, computational curriculum for introductory statistics. Our overarching goal is to equip demographically diverse students with statistical thinking and domain-based skills that are transferable across academia and industry.

Authors

• Anjali Krishnan, Brooklyn College

• Laura A Rabin

• Matthew Crump