ASM events
This conference is managed by the American Society for Microbiology
Table of contents
No headers

As of June, 2009.

1.  Smith, M.K., Wood, W.B., Aedams, W.K. Wieman, C., Knight, J.K. Guild, N., Sue, T.T., (2009).  Why Peer Discussion Improves Student Performance on In-class Concept Questions.  Science 323, 122-124.

This study came out in January of this year and I was very impressed with it.  It looks at the effect of small groups discussing in-class questions.  It asks whether students who initially got a question wrong get it right after group discussion purely due to the influence of more knowledgeable teammates, or because they gained in actual understanding, and they conclude that actual understanding accounts for the improvement in student performance.   Although the vehicle (in-class questions) is different from group exams, both have student discussion with peers in common.  Also, in their methodology the authors used the idea of isomorphic questions - questions that address a concept in similar ways but are different - to look at improvement in understanding over the course of a semester.  Isomorphic questions are likely to be something I will need to use, and their online supplementary material gives a number of examples.

2. Giuliodori, M.J., Lujan, H.L., DiCarlo, S.E., (2008).  Collaborative group testing benefits high- and low-performing students.  Adv. Physiol. Educ. 32, 274-278.  doi: 10.1152/advan.00101.2007.

The authors wanted to test the hypothesis that all students, both strong and weak, benefit from collaborative group testing.  They tested 65 students individually, then immediately after students answered the same questions in groups of two.  The individual and group scores were compared, with students identified as high or low performers based on their individual scores.  They found that group exam scores were significantly higher (mean of 70.2% correct) than individual scores (mean of 58.7% correct).  Furthermore, high performing students improved their scores by a mean of 3%, while low performing students improved by a mean of 20%.   Earlier studies have demonstrated that mean student scores improve with group exams, but this study dissects whether both strong and weak students show improvement, and demonstrates that they do.  Their data provide good arguments to convince strong students about the value of collaborative exams even to them - although the increase might seem small, most students are keen to increase their point totals by even the smallest amount.  The authors also determine the effect size of group exams - effect size is a new concept to me and is something I may need to consider for my own study.

3.  Lusk, M., Conklin, L., (2000).  Collaborative Testing to Promote Learning.  J. Nursing Education 42(3):121-124.

These authors define collaborative testing as “a method of cooperative learning in which students work together but then turn in their own work.”  This is a little bit different wrinkle on the group exam idea, which usually takes the form of a single exam representing group consensus on the answers.  Lusk and Conklin tested for differences in retention between students who tested collaboratively as they defined it, and those who tested individually on exams during the semester by testing them again (individually) on the same concepts during the final exam.  They found no difference on final exam scores between the two groups, and concluded that collaborative testing makes no difference in comprehension or retention.   This study is directly relates to the question I’m asking, and it will be important for me to think about why they found the results they did, especially in light of the next article on my list.

4.  Cortright, R.N., Collins, H.L., Rodenbaugh, D.W., DiCarlo, S.E., (2003).  Student Retention of Course Content is Improved by  Collaborative-Group Testing.  Adv. Physiol. Edu. 27(3), 102-108.

Cortright et al., wanted to know if collaborative testing improved student retention of knowledge, so this study is directly relevant to my own question.  Their design is an example of a randomized crossover study, an experimental design method that is new to me.  They used 2 student groups (Group A and Group B) of 19 each; all were given exam 1 individually, then Group A was given a subset of questions to answer in teams of 2 or 3 students.  A month later, both groups were given exam 2 plus the same subset of questions from exam 1 used for the collaboratively tested group. For exam 2, Group B then received a subset of exam 2 questions to answer collaboratively.  Finally, for exam 3 both groups took the exam individually, with the exam 2 group question subset included.  They found that student retention when asked the same questions later in time was improved when students worked on questions in groups, raising the percent of questions answered correctly from 46.0% (questions answered only individually) to 52.9% (questions discussed in groups), a difference of 6.9%.  They note a limitation of their design is that the students given the group questions have additional exposure time to those questions that the non-group test students did not receive, and one could argue that the improved retention is due to this extended exposure, not to the group discussion of the question.

5.  Kapitanoff, S.H.,(2009).  Collaborative Testing: Cognitive and Interpersonal Processes Related to Enhanced Test Performance.  Active Learning in Higher Education 10(1), 56-70.  doi: 10.1177/1469787408100195.

Kapitanoff was interested to know how collaborative testing works to improve student test scores.  She proposed that it might work in several ways, including through cognitive processes, interpersonal interactions, and reduced test anxiety.  She took a group of 33 participating students and 16 non-participating students and gave them a pre-test questionnaire about test anxiety, confidence, etc. and a week later administered the exam.  All students took the exam as individuals, and immediately following the participators answered a subset of the exam questions in small groups.  Kapitanoff found that students performed better on the group test portion than on their individually taken exam, with a mean gain of 8.4%.  Subsequent to the exam students filled out a follow-up questionnaire.  Kapitanoff bases her conclusions about how group tests improve test performance on what students self-reported about themselves and their attitudes in this follow-up.  Drawing conclusions based on student self-perceptions seems to tell us more about why students think it works than about how it actually does work.


Tag page
You must login to post a comment.