Missing persons: a national survey approach assessing the predictors of case outcomes

ABSTRACT

Using national online surveys of U.S. adults who know a missing person, the current study explores factors influencing missing person case outcomes (i.e. the person being found alive, an arrest occurring). Both demographic characteristics (e.g. the missing person’s sex, race/ethnicity, income) and search activities (e.g. case reported to the police, social media posts) were significantly related to these case outcomes. Multivariate analyses confirmed the strong net effect of federal police reporting on case outcomes, but the impact of search activities and the missing person’s socio-economic characteristics were context specific. These results are discussed in terms of their implications for public policy and future research on the nature and sources of disparities in missing person case outcomes across different situational contexts.

KEYWORDS:

• Missing persons
• search activities
• police reporting
• case outcomes

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. We provide this estimated rate of case solvability because NCIC does not include a direct measure of case outcome or how a particular case was ‘solved’ in its publicly accessible data reports. NCIC only uses the language of ‘entry’, ‘purge’, and ‘end-of-year active missing person counts’ to designate the aggregate flow of cases reported and eliminated from the database. However, because cases may be purged by the reporting agency due to both actual case resolution and administrative/data entry errors (e.g. invalid records, duplicates, clerical error), our derived measure of case outcome from NCIC data is likely to overestimate the rate of case clearance. This is due to the inability in these data to eliminate the unknown number of invalid records from the total derived count of cleared cases.

2. The survey used in the current study asked a question about the respondent’s personal relationship to the missing person, including the following categories: 1) my parent, 2) my sibling, 3) my child, 4) other relative, 5) close friend, 6) acquaintance, and 7) other – please specify. The analyses presented in the following sections use dummy coding of these categories to contrast information provided by family members (#1–4) of the missing person and other non-family (#5–7). To increase the validity of the survey responses, all survey items in this study include the category of ‘unsure/don’t know’ to minimize the impact of forcing respondents to provide specific answers to questions when they lack such information.

3. Given the exploratory nature of the current study, a lower benchmark (p < .10) was used for identifying statistical significance.

4. The lack of significant net effects of search activities and particular case circumstances on the likelihood of being found alive is due in a large part to the significant positive correlation between these variables and the likelihood of reporting cases to the police. Once controls for police reporting are included in these multivariate models, the net effect of both search activities and circumstances become non-significant. As described shortly, contextual analysis is also employed to further explore whether the effect of these independent variables varies or is invariant across different situational contexts.