To the Editor:
The new Institute of Education Sciences-funded study (“Character Education Found to Fall Short in Federal Study,” Oct. 21, 2010) is not the largest federally funded study to date of schoolwide character education programs. In fact, the reported study was a cluster of small studies using convenient invited samples in competitively selected locations that lacked generalizability to larger populations. None of the programs included in the studies was a comprehensive, integrated multidimensional character education process.
To complete a combined analysis of the individual studies is analogous to the Food and Drug Administration commissioning several drug companies to implement their favorite prescription medications to small, convenient samples (with randomized controls) for three years to determine if prescription medications have positive effects on the human condition. No concern for previous condition (or need for medication) is required. One company tests its heart medication on seven subjects from Buffalo, N.Y.; another tests its anti-depression medication on five subjects from North Carolina; and so on.
At the end of the three-year period, the FDA releases a report on the results with the headline “Prescription Medication Found to Fall Short in Federal Study.”
Further examination of the character education study reveals that more than one-third of the students, parents, and teachers did not provide the requested survey and other reporting information, resulting in sampling bias. Add to this that the baseline data were collected in the fall after treatments had begun; one has to wonder about the validity of the data reported. All data analyses were from the first-year fall survey administration to first-, second-, and third-year spring survey administrations. Yet there is a strong research base that says results of student and teacher surveys on school environmental variables decrease from the fall to spring administrations, and that student results drop longitudinally as students increase in age. One would expect negative results from this design, all things remaining equal.
Larger federally funded experimental studies on the impact of character education with both random selection (for generalizability) and random placement of schools have been completed. These studies were funded through the Partnerships in Character Education Program starting in 2002. Two of the completed projects were the Missouri studies. One experimental study involved 64 randomly selected and randomly placed elementary and secondary schools. The other study included 40 randomly selected and randomly placed elementary schools. These studies showed that comprehensive, integrated multi-dimensional character education has an impact on school environment, student pro-social behavior, and test scores. The results of these and other character education studies will be published in the March 2011 issue of the Journal of Moral Education.
Michael W. Corrigan
Educational Psychology and Human Development
Huntington, W. Va.
Jon C. Marshall
Rapid City, S.D.
A version of this article appeared in the November 10, 2010 edition of Education Week as Character Education Studies Need to Be Comprehensive