DOI: 10.1111/j.1939-3938.2009.01032.x

,  

,  

PMID: 19222691

Issn Print: 1556-3693

Publication Date: 2009/03/01


Print

Measurement challenges in forensic nursing research

Alison M. Colbert

+ Author Information

Author Information: Graduate Forensic Nursing Program, School of Nursing, Duquesne University, Pittsburgh, Pennsylvania

    loading  Checking for direct PDF access through Ovid

Excerpt

As a community of professionals, forensic nurses are in dire need of rigorous scientific research, grounded in valid evidence, supporting the clinical work of forensic nurses all over the world. Such research would focus on vulnerable populations including victims of violence, children and adults at especially high risk for victimization, and incarcerated individuals. But as we know, designing research interventions within these specialized groups necessitates that we carefully consider, in advance, issues from Institutional Review Board approval to the reporting of results. I offer the following observations specific to measurement in an effort to open discussion about how we can, as researchers, think proactively about some of the potential challenges unique to the vulnerable populations we seek to serve.
Measurement is a complicated issue, made more so by technical considerations and statistics. Because of a lack of research in vulnerable populations, we may lack reliable tools for measurement. In the broadest terms, reliability refers to the consistency of measurement and validity refers to how well a tool measures what it is supposed to measure (McDowell, 2006). Researchers are obliged to ensure that the tools they use demonstrate good reliability and validity and that we are appropriately qualified to administer any tool we choose.
Validity is the issue most relevant here: specifically, how do we know that any particular tool is right for our particular study and population? That means asking very specific questions. Has this tool been used with this population before? Are the variables defined the same way as the original tool developers intended? Quality of life measures are frequently problematic here–can one measure “quality of life” in a recent female victim of crime with the same tool that was designed to be used with “healthy” young men?
Also, researchers may need to take into consideration the unique situation of the individual, such as incarceration or community living. For example, let us posit that one wants to measure health-promoting behavior of women incarcerated at halfway houses. Issues such as exercise, diet, and leisure-time activities are significantly different in community and institutional living, and frequently choices are not made by individuals themselves. Accurate data about health-promoting behaviors would have to include these types of details, and tools may need to be amended (and then retested) to fit the situation. Also, what is important–the participant's current life situation, or his/her “normal” circumstances? For a woman with a recent history of interpersonal violence currently residing in a shelter, researchers need to make sure their tools ask questions appropriate to the investigatory context. Is the subject of study her life in the shelter, or her previous life experience? Does the question make that clear?
Strickland, Dilorio, Coverson, & Nelson (2007) reviewed nursing research in vulnerable populations and found significant problems with measurement and inadequate attention to issues related to measurement. They made several important recommendations to researchers hoping to improve the quality of their studies, such as conducting gender-based reliability and validity, using a variety of approaches to data collection in addition to self-report, and providing more detailed explanations of conceptual definitions of variables. While these are broad considerations, they do offer concrete ways for researchers to both determine reliability and validity of their own tools and contribute to the body of literature for use by others.
Researchers investigating sensitive topics, such as history of interpersonal violence or legal matters, need to be especially aware of how these experiences may influence data collection.
    loading  Loading Related Articles