Case
Abstract
On 11 September 2001, terrorists attacked the World Trade Center in New York City, resulting in the destruction of the twin towers of the complex and a number of surrounding buildings, killing approximately 2800 people. Using random-digit dialing, we surveyed 2368 New York City residents by telephone 1 year after the disaster. To be included in the survey, respondents had to have been living in New York City on the day of the attacks. One year later, we re-interviewed as many of the original respondents as possible (N = 1681). The cooperation rate for the baseline interview was 63%, while the re-interview rate was 71%. We used structural equation modeling to assess the relationships among exposure to the World Trade Center Disaster, perievent panic attack—a panic attack soon after experiencing or learning about the disaster—and later depression. This case study provides a detailed description about the use of structural equation modeling to evaluate how well the data support the researcher's view of the world. We describe the strengths and weaknesses of surveys and structural equation modeling and walk through a specific structural equation modeling model. We pay particular attention to when structural equation modeling is the most appropriate statistical technique to use and when it is not.