Violence and Ownership

Scientists in the social and behavioral sciences deal with many datarelated

obstacles in conducting empirical research. These include lack

of relevant data, data that are error-ridden, and data that are not based

on properly designed statistical samples (i.e., are unrepresentative) of the

targeted population. These obstacles are particularly difficult in firearms

research. In firearms and violence research, the outcomes of interest, although

large in absolute numbers, are statistically rare events that are not

observed with great frequency, if at all, in many ongoing national probability

samples. Moreover, response problems are thought to be particularly

severe in surveys of firearms ownership and violence. In the committee’s

view, the major scientific obstacle for advancing the body of research and

further developing credible empirical research to inform policy on firearms

is the lack of reliable and valid data.

This chapter summarizes some of the key data collection systems used

to assess firearms policies, describes some of the key properties of useful

research data, and offers some suggestions for how to begin to develop data

that can answer the basic policy questions. There are no easy solutions to

resolving the existing data-related problems. Often, we find that the existing

data are insufficient, but how and whether to develop alternative data

sources remain open questions. For these reasons the committee urges a

significant increase in methodological work on measurement in the area of

firearms ownership and violence.

The committee does not wish to paint an overly pessimistic picture of

this research area. The existing body of research, as described in the other

chapters of the report, has shed light on some of the most fundamental

questions related to firearms and violence. However, in key data areas—the

availability of firearms, the use of firearms, and the role of firearms in

injuries and death—critical information is absent.