Evaluating Research Methods

By James Matthews

Scenario 1: Ten students are available for in-depth interviews. Participants will be selected based on their involvement with the peer mediation program. They will be observed over three weeks. Analysis will attempt to determine issues concerning peer mediation.

I would classify this situation as a Phenomenological study. In this type of study the researcher “collects data on how individuals make sense out of a particular experience or situation” (McMillan & Schumacher, 2006, p. 26). The in-depth interviews following the 3-week observation will offer the requisite of access.


Scenario 2: Two classrooms of students are selected. There are 30 students in each class; each group will have similar demographics—age, sex, race, socio-economic background, etc. Classes will be randomly divided into two groups of 15 students. Of these two groups, one randomly selected group will get training on peer mediation and the other group will not. Thus in each classroom there will be one group that is trained in peer mediation and one that is not. Analysis will occur on which groups have the fewest office referrals.

I would classify this as a True Experiment because of the random assignment of groups. In general, the goal of these types of experiments is “to investigate cause –and-effect relationships between interventions and measured outcomes” (McMillan & Schumacher, 2006, p. 24). In this specific case, the application of the stimulus of training on one group and not the other is the basis of the analysis.


Scenario 3: A school counselor is interested in knowing how student attitudes affect the value of peer mediation to decrease the number of office referrals that are being filed for inappropriate interactions.

This is an Ex Post Facto experimental design. A counselor is trying to find a possible relationship between the number of office referrals and the existence of peer mediation. Further, the “possible causes are studied after they have occurred” (McMillan & Schumacher, 2006, p. 25).


Scenario 4: Peer mediation has become widely used in many schools. The feelings of those involved in the process are little known—either from those doing the mediation or those receiving it. The ZASK-R Acceptance Preference Survey will be given as pre- and post-tests to 40 students participating in mediation. Follow-up interviews will be conducted on a bi-monthly basis.

This is a Mixed-Method Explanatory design since it will collect quantitative data first and then gather subsequent data to “elucidate, elaborate on, or explain the quantitative findings” (McMillan & Schumacher, 2006, p. 28). Specifically, a sample of students will be given surveys in order to try to determine their collective thoughts on the effectiveness of peer mediation. There will also be follow-up in subsequent months.


References:
McMillan, James H & Schumacher, Sally (2006). Research in education: Evidence-based inquiry. Boston, Massachusetts: Allyn and Bacon.