• By Dartington SRU
  • Posted on Friday 23rd March, 2012

Exorcising the ghosts of family intervention research

Nearly a century ago, a Chicago factory called the Hawthorne Works decided to find out what light levels made their workers most productive. As an experiment, it was a disaster. When experimenters turned the lights up, productivity went up. And when the experimenters turned the lights down, productivity went up – again.Later scholars concluded that the “Hawthorne effect” came about because the workers enjoyed the attention of the experimenters. For modern-day scientists, the “Hawthorne effect” and the problem of attrition have become two of prevention research’s most perplexing ghosts. Now, researchers on a teen and family intervention claim to have found ways to exorcise both phantoms. The Rural African American Families Health (RAAFH) project is a randomized trial evaluating the effectiveness of the Strong African American Families–Teen (SAAF–T) program, a family-centered prevention program for rural African American adolescents. RAAFH claims to be the first family-centered prevention trial to use a design that rules out “Hawthorne effects” – and it also retained twice as many families as similar programs. The GhostsWhy exactly are these two ghosts a problem for research? First, the “Hawthorne effect” is the idea that drawing attention to an issue is enough to change it. What makes this a threat is that the researcher cannot be sure that improvements result from the “active ingredients” of an intervention. The particular risk for RAAHF is that simply bringing families together in a supportive environment, rather than the actual content of the SAAF-T program, would produce change. Second, “attrition” – when participants leave a study before it ends – is a threat because there may be something unique about the people who stayed in the study that made them more likely than those who left to change their behavior. The particular attrition risk for RAAHF is that high attrition rates are common with universal, family-centered programs in general, but are even more so for RAAFH’s target population of rural African-Americans. The ExorcismThere are two components of the RAAFH prevention trial that make it uniquely able to exorcise these ghosts, claim the researchers, led by Steven Kogan of the University of Georgia. First is its attention-control design to rule out the Hawthorne effect, and second is an extensive engagement protocol used to minimize attrition. In an attention-control trial, the study participants receive either the intervention being tested or an intervention that is similar in structure, but quite different in content. The comparison program will be structured and delivered in a method exactly the same as the tested program, but will be designed to produce very different effects. The SAAF-T program was designed to enhance protective family management processes with the aim of reducing substance use and risky sexual activity. To create an attention-control for the SAAF-T program, the researchers used a health promotion program that specifically did not include content about substance use or sexual risk reduction and did not target family relationships. Instead, they designed a comparison intervention to promote healthful behaviors among adolescents by encouraging good nutrition, exercise, and informed consumer behavior. The two programs were identical in duration and format. Both used videotaped presentations and interactive components to facilitate learning. The point is that the control group was set to receive the same amount of attention as the intervention group – so that any differences could truly be attributed to the intervention content.To prevent attrition, on the other hand, the extensive engagement protocols included several important components. First, members of the community served as liaisons to help recruit families to the study. They were selected on the basis of their positive reputations and extensive social contacts in their communities. They were trained on the goals of the project and, according to Kogan and colleagues, provided a credible community voice to establish trust between families and the research team. After recruited families were randomly assigned, they received an information visit by a program leader who explained the program and answered questions from the family. An important feature of RAAFH is that the program leaders were African Americans from the local community. The program leaders conducted the information visits and delivered the intervention programs. By having community members serve as program leaders, RAAHF gained in credibility and acceptability, the researchers said. Finally, extensive practical support was provided to encourage family attendance. Support included transportation, childcare for younger siblings, payment for attending sessions and completing assessments. The families also shared a meal together before each session, which not only encouraged attendance, but also promoted family bonding. The ResultsRAAFH was successful in exorcising its research ghosts, Kogan and colleagues claim. Checks of the RAAFH program implementation indicate that both SAAF-T and the comparison program were delivered as intended and that, on average, the families in each group received the same number of sessions. Importantly, at the end of the intervention period, the SAAF-T youth had higher levels of protective family management skills than the comparison youth. The effective implementation of the study design ruled out “Hawthorne effects” or the use of structured family interactions as explanations of these results.RAAFH protocols were equally successful in minimizing the threat of attrition. Among all pre-tested families, only 7% left the study before attending at least one intervention session. Only 12% of the families left the study before attending at least four of the five program sessions. These rates are similar to the lowest attrition rates obtained in family-centered prevention trials with middle-class Caucasian parents. In the end, 62% of all families attended the full five sessions - a rate, RAAFH researchers claim, that is more than twice that from trials of other family-centered programs. ********Reference:Kogan, S.M., Brody, G.H., Molgaard, V.K., Grange, C.M., Oliver, D.A.H., Anderson, T.N., DiClemente, R.J., Wingood, G.M., Chen, Y. & Sperr, M.C. (2011). The Strong African American Families-Teen Trial: Rationale, Design, Engagement Processes, and Family-Specific Effects. Prevention Science. DOI 10.1007/s11121-011-0257-y.

Back to Archives