Reconsidering Homicide Clearance Research: The Utility of Multifaceted Data Collection Approaches

• DOI: 10.1177/1088767920939617
• Author: Brent Teasdale,Shila René Hawk,Dean A. Dabney
• Published date: 01 August 2021
• Date: 01 August 2021
• Subject Matter: Articles

https://doi.org/10.1177/1088767920939617

Homicide Studies

2021, Vol. 25(3) 195 –219

© 2020 SAGE Publications

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1088767920939617

journals.sagepub.com/home/hsx

Article

Reconsidering Homicide

Clearance Research: The

Utility of Multifaceted Data

Collection Approaches

Shila René Hawk1, Dean A. Dabney2,

and Brent Teasdale3

Abstract

This study explores issues associated with the data commonly used in homicide

clearance research. Data collected from 2009 to 2011 case files (n = 252) were

reviewed during interviews with investigators (n = 29). The multifaceted data

collection approach produced a more comprehensive dataset than was available based

solely upon case file reviews, with alterations to the data occurring in as many as 69%

of the cases. The process advanced the precision of the data recorded, reduced

missingness, and heightened detail on key variables. Significant differences were noted

in multivariate analyses of the datasets when modeling clearances. Findings suggest

contextualizing case file data is valuable.

Keywords

homicide, clearance, policing, multimethod, missing data, investigation, methodology,

solvability

Widely recognized as the most thoroughly documented crime by law enforcement,

criminologists have long turned to homicide data to inform the theoretical- and policy

landscape of the field. Simply stated, no other type of offense more centrally informs

criminology than homicide. Consider the following observations in this regard: 1)

nearly the entire subfield of macro criminology is built around studies using city, state,

1Applied Research Services, Inc., USA

2Georgia State University, USA

3Illinois State University, USA

Corresponding Author:

Shila René Hawk, Applied Research Services, Inc., 3235 Cains Hill PL NW, Atlanta, GA 30305, USA.

Email: shawk@ars-corp.com

939617HSXXXX10.1177/1088767920939617Homicide StudiesHawk et al.

research-article2020

196 Homicide Studies 25(3)

or national homicide data; 2) rarely does a year pass that top-tier journals do not pub-

lish multiple articles built around homicide data; 3) the journal Homicide Studies pub-

lishes four issues annually dedicated exclusively to the phenomenon; and 4) the

National Institute of Justice maintains a robust portfolio of funding, research reports,

and publicly available databases focused on homicide and its prevention. Given its

prominence in our scholarly and public thinking about crime, it is critical that our

understanding of the occurrence of and response to homicide be shaped by the most

valid and reliable data possible.

This study employs a series of homicide data enhancement protocols and empirically

examines the degree to which these efforts impact the prediction of homicide case clo-

sure. There exists considerable scholarship on the topic of homicide case closure that

spans many decades (see Riedel, 2008 and Hough & McCorkle, 2020 for a general

overview). Still the homicide clearance literature needs increased methodological con-

versation to help remediate the fact that findings are mixed, incongruent, and even con-

tradictory (Lee, 2005; Riedel, 2008; Wellford & Cronin, 1999). Many research questions

remain unanswered (Jarvis & Regoeczi, 2009; Jiao, 2007; Mancik et al., 2018; Rydberg

& Pizarro, 2014). Arguably the next step in advancing the homicide clearance literature

is for research to move past the data limitations of previous studies.

Scholars have noted that clearance studies based on archival data suffer from qual-

ity and completeness issues that significantly compromise how they can be causally

interpreted (Alderden & Lavery, 2007; Lundman & Myers, 2012; Puckett & Lundman,

2003; Riedel & Rinehart, 1996). We investigate data quality as a potential source of

these issues. We observe that most clearance research relies exclusively on archival

data and those data sources may not accurately reflect the facts of the criminal event

and subsequent investigation efforts. Research using these data are often described as

having imprecise measures or omitted factors that likely shape case outcomes. These

data quality issues may pose a threat to models seeking to account for the factors that

predict case closure and the theoretical- and policy-related insights that follow.

When relying on agency-released data, homicide researchers have limited access to

a predefined set of measures and issues of missing data (Riedel & Regoeczi, 2004).

This results in gaps in the inclusion of key measures, the use of proxies, and compro-

mised reliability (Pizarro & Zeoli, 2013; Roberts, 2007). Similar problems are

observed in comprehensive archival datasets assembled directly from official case

files, as they are generally restricted by agency record-keeping practices and lack

some key information among ongoing investigations (Rydberg & Pizarro, 2014;

Schroeder & White, 2009). This means that data concerning relevant case characteris-

tics and investigation activities are largely absent (Carter, 2013; Keel et al., 2009;

Regoeczi & Jarvis, 2013). Furthermore, these key data points are often systematically

missing, given that homicide investigation narratives generally constitute truncated

and sterilized prosecution-oriented documents as opposed to a totalistic account of the

investigation process (Hawk & Dabney, 2019; Innes, 2003).

The current study systematically compares contemporary homicide case file data

augmented through interviews with the investigators who constructed the files. We

examine if the above-mentioned precision and robustness problems could be