Study: When Helping Hurts (Research Methods)
Many years ago, researchers conducted one of the most ambitious studies designed to investigate whether we could prevent juvenile delinquency in at-risk youths. The research project was funded by a retired professor of social ethics and medicine named Richard Cabot. Dr. Cabot believed that children and adolescents develop behavior problems because they lack prosocial role models in their lives. He believed that at-risk youths could be “steered away from delinquency if a devoted individual outside his own family gives him consistent emotional support, friendship, and guidance” (McCord, 2010, p. 33).
To test this hypothesis, researchers recruited over 500 boys from Cambridge and Somerville, two densely populated, low-income neighborhoods near Boston. Many of the boys were identified by teachers as exhibiting conduct problems. At-risk boys without behavior problems were also recruited for the study to prevent stigma associated with participation.
The researchers matched boys based on certain characteristics that might predict their likelihood of delinquency, such as their intelligence, level of aggression, home environment, and parents’ use of discipline. Then, the researchers randomly assigned one boy from each pair to a treatment group and the other boy to a control group.
Each boy in the treatment group was assigned a social worker who served as his mentor. Mentors met with each boy approximately twice every month: helping him with his homework, taking him to the YMCA or other activities, assisting him in finding a job, or engaging in sports or games. Mentors also worked with boys’ families to help them improve the quality of family life. Boys in the treatment group were also allowed to attend a free summer camp with other boys in the program.
Figure 1 ■ Outcomes of the Cambridge–Somerville Youth Study
Note: Although therapeutic summer camps were designed to help at-risk youths, boys who attended these camps multiple times were more likely than other boys to have unfavorable outcomes later in life (Dishion et al., 1999).
Boys in the control group received no special assistance, although they could seek treatment as usual in the community. The Cambridge-Somerville Youth Study lasted 5 years and was the first large-scale randomized controlled study investigating the effects of a psychosocial intervention on children.
Initial results were disappointing. Despite 5 years of mentoring, boys in the treatment group did not show better outcomes than the controls. Nevertheless, researchers remained optimistic that boys who received the treatment would fare better as adults.
Thirty years later, Joan McCord examined the long-term effects of the program. Amazingly, she was able to track down 98% of the boys who participated in the study. (They were now middle-aged men.) Then, she analyzed public databases, hospital logs, and court records to examine the boys’ long-term outcomes.
Most of the boys who participated in the treatment recalled the program favorably. Similarly, most mentors believed that treatment had a positive impact on the boys’ lives. However, when McCord examined the objective data, she was shocked. “It was so surprising,” she said. “Everything was going in the wrong direction” (Zane, Welsh, & Zimmerman, 2017, p. 152).
McCord discovered that boys who received mentoring were more likely than controls to commit a felony, to develop substance use problems, or to manifest serious mental illness. Boys in the treatment group were also more likely than controls to have stress-related health problems, to report low job satisfaction, or to die at an earlier age. Indeed, 42% of boys who received mentoring experienced at least one of these “undesirable outcomes” compared to 32% of boys in the control group. The Cambridge-Somerville Youth Study backfired; it increased the likelihood of behavior problems in children who received treatment. What went wrong?
McCord and colleagues discovered that boys who spent more unstructured time with other boys in the program were at greatest risk for having negative outcomes later in life. In particular, boys who attended summer camp once during the treatment program were only 1.3 times more likely than controls to develop conduct problems. In contrast, attending camp multiple times increased a boy’s likelihood of developing conduct problems 10-fold (Figure 1). The researchers concluded that boys talked about and reinforced each other’s delinquent behavior during camp. The more time they spent at camp, the greater their likelihood of antisocial behavior later in life (Dishion, McCord, & Poulin, 1999).
Discussion Questions
1 How does the Cambridge-Somerville Youth Study illustrate the scientific principle of critical thinking?
2 Which of the four goals of psychological research did the Cambridge-Somerville Youth Study try to achieve?
3 Was the Cambridge-Somerville Youth Study’s design cross-sectional or longitudinal?
4 The researchers discovered that the effects of treatment on children’s outcomes depended on how many times they attended the summer camp. Boys who did not attend or attended the camp 1 time usually did not show undesirable outcomes, whereas boys who attended 2 or more times were at greater risk for undesirable outcomes. In this study, does “summer camp attendance” serve as a mediator or moderator variable?
5 What is the difference between random selection and random assignment? Which did the researchers use in the Cambridge-Somerville Youth Study?
6 What were the independent and dependent variables in the Cambridge-Somerville Youth Study?
7 What type of control group did the researchers use? What was one potential benefit of this type of control group?
8 What is attrition? How can it threaten the internal validity of a study? Was it a problem in the Cambridge-Somerville Youth Study?
9 What is external validity? In what way is the external validity of the Cambridge-Somerville Youth Study limited?
10 What can the Cambridge-Somerville Youth Study teach us about the causes of children’s disorders?
11 What can the Cambridge-Somerville Youth Study teach us about the importance of psychological research, especially regarding the treatment of child and adolescent problems?
References
Dishion, T. J., McCord, J., & Poulin, F. (1999). When interventions harm. American Psychologist, 54, 755–764.
McCord, J. (2010). The Cambridge-Somerville study. In G. Sayre-McCord (Ed.), Crime and family (pp. 32–40). Philadelphia, PA: Temple University Press.
Zane, S. N., Welsh, B. C., & Zimmerman, G. M. (2017). Examining the iatrogenic effects of the Cambridge-Somerville Youth Study. British Journal of Criminology, 56, 141–160.
This case study accompanies the textbook: Weis, R. (2021). Introduction to abnormal child and adolescent psychology (4th ed.). Thousand Oaks, CA: Sage. Answers appear in the online instructor resources. Visit https://sagepub.com.
Descriptions of Images and Figures
Back to Figure
In the graph, x-axis shows the times attended summer camp as never, once, and twice plus, and y-axis shows the odds of an unfavourable outcome from 0 to 10 in increments of 2. The graph shows the following
