Researchers Dana Haynie and Scott South, in their study “Residential Mobility and Adolescent Violence,” published in the September, 2995 edition of Social Forces, provide an interesting look at the relationship between the amount in which a child and his or her family has changed locations and the extent to which this child engages in violent behaviors. These researchers also explore the mediating effects of four variables, including the parent-child relationship, psychological distress, victimization, and peer social networks, and how these variables influence the relationship between mobility and violence. Through a collection of more than 8,000 datasets, Haynie and Scott provide an extensive sample of this relationship, delivering a valuable research contribution to the social sciences. Although their study was generally well conducted and free of research errors, it is not without its faults and the results should be taken with some degree of caution.
The purpose of the current study was to investigate the relationship between residential mobility and child violence. While the researchers clearly state their purpose, they fail to clearly outline their specific hypotheses. While the reader is likely to infer that a higher degree of residential mobility is positively correlated with violent behaviors, the paper neglects including this information in the methods section. Similarly, the researchers do not provide hypotheses regarding their mediating variables, and the reader is left to infer this information on his own.
The measurement is generally sound, as the researchers explain that they conduct a large multivariate analysis of their proposed relationships. Their sampling was not truly random, however, as a self-selected sample of individuals who completed all three series of interviews was used in the final data analysis. Those who did not complete the full three components of the interview process may have dropped out for reasons which would have changed the results of the study, such as the children not wanting to answer personal questions about their violence, or disruptions in the parent-child relationship that could potentially confound the data.
A significant flaw in the methods section was the arbitrary definition of residential mobility. Researchers defined this term as the amount to which the child or his family has changed locations over the past two years. However, the authors failed to cite research explaining that the past two years had demonstrated any influence on child violence, or why this period was important or significant to a child’s psychological state. The authors failed to include a control variable regarding whether the child moved to a situation in which their life improved or regressed emotionally. This arbitrary definition could potentially confound the study. On the other hand, the authors did an adequate job of providing reliability and validity statistics for the mediating variables. Researchers also cited previous research using these measures in the past.
Results demonstrated that children who had experienced more residential mobility over the past two years experienced more violent behaviors than those who did not. Results also demonstrated that the mediating variables significantly influenced this relationship. However, it’s important to note, as author Russel Schutt explains, that correlation does not equal causation. A relationship between variables does not explain the relationship or reveal that one variable led to another. Similarly, because of the lack of reliability or validity statistics on the residential mobility measure, it’s difficult to generalize this study to the general population. This study used a self-selecting sample, and may have neglected to include students who experienced the most severe violent behaviors or those who did not complete the study because they did not have a history of violence.
While the study may lack some degree of generalizability, it appears to be logical and explains a crucial piece of understanding child violence. It is logical to believe that residential mobility may influence violence, as children who have a more tumultuous home life are likely to use violence as an outlet. The internal validity of this study is sound, but the external validity is somewhat lacking because of the sample type. This study only applies to American children from the ages of 7th to 12th grade. However, the study was detailed in controlling for demographic variables such as sex, race, and socioeconomic status to increase the external validity within this population.
In conclusion, Haynie and South make a valuable contribution to the social science literature with their exploration of the relationship between residential mobility and violence. Results demonstrating that high levels of mobility are related to violence can help researchers and professionals in the social sciences to understand violence and perhaps intervene when children must move from one high school to another. However, because of the lack of causation, the possible self-selecting sample, and the limited validity in some cases, results from this study must be taken with some degree of caution. Future studies should use a true random sample, and should employ a non-correlational design to obtain more causal results.
Haynie, Dana, and South, Scott. “Residential Mobility and Adolescent Violence.” Social Forces. 84.1 (2005): 361-374.
Schutt, Russell. Investigating the Social World: The Process and Practice of Research (3rd Ed.). Thousand Oaks, CA: Pine Forge Press, 2001.