Using the exhibited generalization approach to evaluate a carbon monoxide alarm ordinance.

Current interests in enhancing the focus of external validity or transferability call for developing practical evaluation approaches and illustrating their applications in this area for meeting the need. This study takes the challenge by introducing an innovative evaluation approach, named the exhibited generalization approach, and applying it in evaluating the carbon monoxide (CO) alarm ordinance passed by Mecklenburg County, North Carolina. The stakeholders specifically asked evaluators to determine the answers to the following two questions: (1) Does the alarm ordinance work? (2) What generalizable information can the Mecklenburg experience provide to other jurisdictions trying to decide if the alarm ordinance's planning, implementation, adoption, and outcomes are transferable to their communities? This study illustrates how to apply the exhibited generalization approach to provide the stakeholders with answers to these questions. Our results indicate that the alarm ordinance was effective in increasing CO alarm ownerships and reducing CO poisoning cases. The evaluation provides potential users and other interested parties with the necessary information on contextual factors and the causal mechanism underlying the CO alarm ordinance, so that these parties and users could decide whether the Mecklenburg alarm ordinance would be transferable to their own communities. Discussions include implications of this study for contributing in further advancing evaluation theory in addressing transferability or external validity issues.