The PLTL Boost

The PLTL Boost: A Critical Review of Research is an article in Progressions that summarizes the body of research studying the impact of PLTL on student success in STEM courses, and contributes a critique of statistical models that has a wide implications for future studies of student performance. The main points are:

1. There are 15 published studies that represent introductory college science and engineering courses at 23 institutions that show an average increase of +15 % ABC for students participating in PLTL. 
2. Three main types of implementations were identified: Type 1: a one to two hour peer-led workshop that replaces one lecture; Type 2: a one to two hour peer-led workshop that replaces a graduate student led recitation; and Type 3: an optional, added peer-led led workshop. All three types of implementation demonstrated similar impacts for participating students. 
3. The studies used a variety of statistical models, from simply assuming a random distribution to more sophisticated multi-linear regression. It is shown that multi-linear regression has an inherent error when attempting to control for prior knowledge in the guise of such things as SAT scores. The result is that the impact of PLTL is underestimated for many students. An alternative, more robust model, is proposed: comparative linear regression, as describe in Zar (Biostatistical Analysis). 

If we adhere to the principles of Scientific Teaching, then the large body of work reviewed here should provide compelling evidence for those interested in student success to closely examine this straightforward adaptation of traditional teaching models.