Overcoming the fear of statistics: Integrating statistics within a research methods course
By Nicholas A. Turiano, PhD
Incoming freshmen are excited about many things they will get to experience for their first time during their transition to college: newfound independence from their parents, meeting new friends, exciting new classes and a long list of other life changing experiences. Unfortunately, taking a statistics course is not high on that list. Instead, taking a statistics course is accompanied by feelings of anxiety, dread and even panic. Likewise, faculty members experience almost the exact emotions when they are told they have been chosen to teach these large-lecture statistics courses.
This cycle of negativity needs to be broken so undergraduate students can successfully gain skills in statistics. A way to lessen the fear of statistics and to ensure the successful transfer of statistics knowledge to the real-world is to teach an integrated two-semester course that combines both statistics and research methods.
The common instructional model in the social sciences is for students to take an introductory statistics course and then a separate research methods course. Several problems exist with this model. First, students often struggle in these large lectures early in their college career, so a substantial number of students do not pass the statistics course and need to repeat it. Because a basic statistics course is often a prerequisite for a research methods course and many other upper level courses, failure to complete the course can delay student progress. Second, the statistics course is usually taken outside of social sciences within a statistics or math department. This is a problem because these courses are taught for students across the entire university and the content is sometimes at a level that is not useful for someone in the social sciences.
Any statistics training is beneficial, but when material is not taught in a manner that is directly applicable to the field students are studying, there is no real motivation to learn or retain such information. For example, why do students need to know how to analyze and interpret a t-test if they are studying to be a clinical psychologist? There are plenty of reasons why a clinical psychologist would need to know this information, but statisticians are not well versed in applied scenarios. The application of statistical methods is what we want for our students, but that is seldom achieved when statistics are taught without the context of the student’s own discipline. Within the current model, even if students are able to pass the statistics class and move on to an upper level course, that does not necessarily mean they have learned the skills needed for those classes. Many faculty spend time reteaching the statistics students should know by the time they enter a research methods course. It is a satisfying feeling when the lightbulb goes off in students’ minds as they exclaim that they finally get it, but it takes time away from learning other material. After a few semesters of the same happening in my own research methods course, our department decided it was time to start teaching statistics in-house by creating a two-semester integrated statistics and research methods course.
The idea of such a course came from recent research from Barron and Apple (2014). The authors provided empirical evidence that an integrated course (compared to nonintegrated) not only improved learning in the course, but also improved scores on the statistics section of the Psychology Area Concentration Achievement Test (PACAT). Utilizing suggestions from the article, we designed a two-semester sequence where students have a large lecture to learn principles of research design and different types of statistical procedures. Lecture is followed by a weekly lab broken into smaller sections so graduate student assistants can directly apply what students learned from lecture to real-world studies. In the first semester, students learn how to measure psychological variables, how to design descriptive and correlational studies and how to use SPSS statistical software to assess central tendency, variability and compute correlation coefficients. In the second semester, students build on these skills and learn about reliability and validity and how to conduct quasi-experimental, experimental and single-subject research designs. Students then learn how to use SPSS to compute t-tests, ANOVA, interactions and a primer in linear regression. Within both courses, the students conduct their own group research projects, write a full APA style research paper and present their findings to the class.
Two key aspects of a successful integrated course are to have hands-on applied experience in smaller lab sections and to have graduate teaching assistants who can walk students through aspects of research design and statistical analyses. For example, it is helpful to have live demonstrations of experiments in which the students are the participants. They are directly exposed to random assignment and experimental manipulations, data are collected and input live into an SPSS shell and a t-test is conducted and interpreted all in real-time. This direct application of the design and resulting statistical method in the lab setting is what students need to truly understand how to conduct quality research and statistical procedures. Students feel confident in their abilities if instructors can describe the important estimates from SPSS output and how it should be interpreted in regard to the research question.
Overall, there are many benefits of offering a two-semester sequence that integrates research methods and statistics. Not only does it aid students in the acquisition and retention of skills, but it creates a natural building of skills over time so students are not thrust into an environment where they are required to cram as much statistics in their minds only to wait a semester to figure out how that applies to psychology. It also promotes effective writing as many students first learn empirical APA style writing in their research methods course. A two-semester course gives students a double-dose of writing feedback that is welcomed by faculty teaching upper level courses. Not all institutions will have the opportunity to offer statistics in-house, but instructors can at least attempt to better integrate statistics into their research methods courses because that is what is needed to lessen the fear of statistics.
- First semester: Descriptive and Correlational Research.
- Second semester: Experimental and Quasi-Experimental Designs.
Barron, K.E., & Apple, K.J. (2014). Debating curricular strategies for teaching statistics and research methods: What does the current evidence suggest? Teaching of Psychology, 41, 187-194.
About the author
Nicholas A. Turiano, PhD, is an assistant professor of lifespan developmental psychology at West Virginia University. His research interests include the study of the mechanisms connecting personality to health and longevity.