Moderators do not account for the relationships between predictor and outcome variables, but rather, they influence the nature of the relationship between the predictor and outcome variables (Breitborde et al., 2010). The moderator model is similar in principal to interaction effects in statistics, where the impact of one variable depends on the level of another variable. Thus the impact of the intervention depends on the level of the moderator variable. A moderator variable could be a qualitative or quantitative variable that can affect the direction and/or strength of the impact that the intervention has on the outcome variable. Again, using an iPad literacy intervention app as an example, a moderating variable could be the clinician’s level of expertise with the app.
In the IES-NCSER Request for Application Document (as well as on some of the IES webpages we viewed), mediator and moderator variables were mentioned as important factors to consider in Exploration and Efficacy and Replication projects. The Request for Application Document highlights the value of exploring mediator and moderator variables below:
“ Mediating and moderating variables. Observational, survey, or qualitative methodologies are encouraged as a complement to experimental methodologies to assist in the identification of factors that may explain the effectiveness or ineffectiveness of the intervention. Mediating and moderating variables that are measured in the intervention condition that are also likely to affect outcomes in the comparison condition should be measured in the comparison condition (e.g., student time-on-task, teacher experience/time in position).
The evaluation should be designed to account for sources of variation in outcomes across settings (i.e., to account for what might otherwise be part of the error variance). Applicants should provide a theoretical rationale to justify the inclusion (or exclusion) of factors/variables in the design of the evaluation that have been found to affect the success of education programs (e.g., teacher experience, fidelity of implementation, characteristics of the student population). The research should demonstrate the conditions and critical variables that affect the success of a given intervention. The most scalable interventions are those that can produce the desired effects across a range of education contexts. “ (IES, 2007).
Breitborde, N.J.K., Srihari, V. H., Pollard, J. M., Addington, D. N., & Woods, S. W. (2010). Mediators and moderators in early intervention research.Early Intervention Psychiatry, 4(2), 143 – 152.
Institute of Education Sciences – Special Education Research Grants. (2007). Request for application [word document]. Retrieved from ies.ed.gov/ncser/doc/2008324_secondary.doc