A Meta-Evaluation of 11 School-Based Smoking Prevention Programs
Tingle, Lynne R., DeSimone, Marie, Covington, Benjamin, Journal of School Health
Use of tobacco products among adolescents has generated much concern in the United States. By age 18, most people who will ever smoke already have done so or are in the process of developing the habit. (1) Therefore, if we can prevent teens from starting to smoke, we will reduce long-term smoking rates. Despite the preponderance of evidence demonstrating the dangers of tobacco use, young people continue to begin smoking at the alarming rate of 3,000 adolescents per day. (2,4) One method of primary prevention involves focusing on schools, sometimes beginning at the elementary level, for smoking prevention interventions. (1)
Research has been conducted including meta-analyses, primary research, and secondary research focusing on the essential elements of smoking prevention interventions, including: 1) information about social influences including media, peer, and parents, (1,5) 2) information about short-term physiological effects of tobacco use, (1) and 3) training in refusal skills. (1,3) Yet, no research has focused on the quality of evaluations of these programs. This study applied the technique of meta-evaluation to determine the level of quality for school-based smoking prevention program evaluations.
The concept of meta-evaluation has gained recognition the past three decades. Formal works on meta-evaluation procedures, guidelines, and checklists emerged in the late 1960s and early 1970s. (6-8) A meta-evaluation involves the process of delineating, obtaining, and applying judgmental information about the utility, feasibility, propriety, and accuracy of an evaluation to report its strengths and weaknesses. (9) Meta-evaluation provides a measure of accountability and fortifies one's ability to judge the merit of research presented. Meta-evaluation also enables researchers to assess the quality of their own work to provide direction and feedback, as well as a sense of professional credibility. (10)
The role of meta-evaluation in research has grown from a beneficial idea to more of an imperative. Stufflebeam (10) deemed it a "professional obligation" of evaluators to ensure they evaluate their work so outcomes from the research will be strengthened. With respect to fields with social service programs created primarily on the strength of research, such as youth smoking prevention programs, meta-evaluations help ensure that the proper questions were addressed, that correct resources were aptly obtained and employed, that measures to analyze data were appropriate, that results from evaluations accurately reflected the efficacy and usefulness of the programs, and that program funds were applied appropriately.
Literature searches were conducted to locate publications concerning school-based smoking prevention programs and meta-evaluations. The main search engines used were ERIC and Medline. Articles and books also were cross-referenced to locate papers possibly missed during the computer searches.
The rating system for this meta-evaluation was created from a combination of examples from past meta-evaluations in other subjects. (9,11,12) The Joint Committee on Standards for Educational Evaluation (13) set four guiding principles for evaluators: utility, feasibility, propriety, and accuracy. Methodology for this meta-evaluation focused on the accuracy component of the four guiding principles. Using examples from previous studies, the authors set criteria for the rating system. The criteria included quality of the research design, estimates of reliability and validity, statistical analyses and interpretations, discussions of practical significance or effects size, attrition effects, and implementation tracking. Using the aforementioned criteria, judges rated each project based on a 0-2 scale, with 0 = criterion was not met, 1 = criterion was met, but not exceptionally well, and 2 = criterion was highly acceptable.
Program evaluations should be designed to minimize threats to internal and external validity. …