Moving toward the Scholarship of Teaching and Learning: The Classroom Can Be a Lab, Too!

Abstract

Division 2 of the American Psychological Association was practicing the scholarship of teaching long before Boyer (1990) coined the term, but Boyer brought the idea to academe's consciousness. Over time, however, the scholarship of teaching and “scholarly teaching” have become confused. Based on Richlin's (2001) work, I attempt to resolve this confusion by equating the scholarship of teaching with scholarly work by faculty. In addition, I point out the parallels between assessment and the scholarship of teaching and learning (SoTL) and provide tips for faculty who want to engage in SoTL research but see roadblocks in their way.

Designing SoTL Studies—Part II: Practicality

“…These measurements can NEW DIRECTIONS FOR TEACHING AND LEARNING • DOI: 10.1002/tl be collected at the same or different times. However, in these designs, it is nearly impossible to establish that one variable caused changes in another variable (Shadish, Cook, and Campbell 2002). Because all students are measured on all assessments, this design can be used when one has a single group of students.…”

“…This design has no difficulty with carryover, testing, or instrumentation effects because each student has only one treatment and is assessed one time. However, as mentioned in the last chapter, the main problem is selection bias (Shadish, Cook, and Campbell 2002). The greater the difference between the two groups, the more arguments can be made that these differences between students caused any changes on the assessment and not the treatment.…”

“…Random Assignment Sounds Great, But How Can I Do It? Random assignment of students to treatment and control conditions creates high internal validity for the study (Shadish, Cook, and Campbell 2002). Unfortunately, random assignment does not occur naturally.…”

“…In all options mentioned in the last section for splitting classes, researchers can use random assignment to assign students to the treatment and control conditions. The class can be split by choosing numbers out of a hat, flipping a coin, using a random number table, and so on (Shadish, Cook, and Campbell 2002). Even assigning students by counting "1-2-1-2 .…”

“…In contrast, extensive use of clickers during the semester may demonstrate a detectable effect. Larger effects have greater statistical power (Shadish, Cook, and Campbell 2002). Higher power through testing large effects can equalize other factors that decrease power such as a small sample size.…”