SAFETYLIT WEEKLY UPDATE

We compile citations and summaries of about 400 new articles every week.
RSS Feed

HELP: Tutorials | FAQ
CONTACT US: Contact info

Search Results

Journal Article

Citation

Johnston BD. Inj. Prev. 2014; 20(2): 73.

Copyright

(Copyright © 2014, BMJ Publishing Group)

DOI

10.1136/injuryprev-2014-041224

PMID

24643884

Abstract

Research in injury prevention, by definition, requires collection of data in some form. Thus every investigator is faced with the challenge of identifying data collection instruments and deciding how best to use these. In some cases, a researcher will be asking new questions or measuring a new concept. If so, the researcher will need to decide how to operationalise the concept as a measurable entity. A process of instrument development and validation may be required. For example, how does one measure “risk taking behavior” or “parental supervision?” These are not trivial questions; getting the measures wrong risks the integrity of the entire study.

On the other hand, many projects involve measuring a behavior, attitude, belief or self-reported outcome that others have measured in the past. Assuming that previous researchers have taken the time to develop and validate a data collection tool, why should others in the field, hoping to measure the same entity, start from scratch? If, for example, I want to measure depressive symptoms, I would almost certainly be better off using an established data collection instrument (like the CES-D) to do so, rather than trying to develop a new measure for my particular study.

For a number of reasons, it is generally preferable to use existing data collection instruments when these are available. Certainly, this increases research efficiency. There is no need to undertake laborious instrument development and validation studies if a well-established measurement tool already exists. Such measures are often explicitly based on a theoretical model of belief or behaviour, a fact that increases the rigour of the investigation and may shape thinking about interpretation of the results. In addition, a researcher can use not only the data collection forms (and the constructs they employ) but also data dictionaries, aggregate variables and data reduction techniques previously developed for use with the instrument.

Whilst reuse of well-developed research tools improves efficiency and rigour, it also facilitates data sharing, direct comparison and even meta-analysis of published studies. If two researchers are measuring the same outcome in the same way, it is much easier to compare or combine their findings in a productive manner. How many studies published in this journal alone would benefit from the use of established and accepted measures of home injury hazards? Fall risk in the elderly? Observed and self-reported parental supervision or child restraint use?

The problem, of course, is that we don't always know that such tools exist and–if they do–are typically challenged to obtain a copy for use in our own work. In this digital age, however, it is not surprising that web-based collections of data collection instruments have begun to appear. The National Institutes of Health toolbox offers brief measures assessing cognitive, emotional, motor and sensory function (www.nihtoolbox.org); the US National Cancer Institute has Grid-Enabled Measures for consensus-rated standard measures in oncologic research; the PROMIS project provides metrics for common patient reported outcomes (physical function, pain, distress, etc); and the Shared Data Instrument Library ties validated instruments into the REDCap secure online data resource.

Happily, researchers in injury and violence prevention now have database and library of non-proprietary data collection instruments and survey forms specific to their field. The US National Center for Injury Prevention and Control at the Centers for Disease Control and Prevention (CDC) funded the Society for the Advancement of Violence and Injury Research (SAVIR) to develop a system to facilitate sharing of non-proprietary instruments among the research and practice communities. Thanks to work led by Carol Runyan and David Lawrence, that system now exists.

Housed within the searchable–and freely available–SafetyLit database, the SAVIR Instrument Library provides information about the development and background of the instruments, how the instruments were used to support research published in peer-reviewed articles and reports, comments on the use of the instrument, any problems encountered, and often, the instrument itself. Citations and links to the published articles based on the instrument are also included. Interested readers can explore the instrument collection here: http://www.safetylit.org/instruments.htm

The goal now is to expend the number of data collection tools in the library. To that end, the journal encourages our colleagues to make use of the collection when launching new research and to contribute their own measures to the database. Study protocols, in particular, should be submitted with data collection instruments when available. We will also be inviting authors of accepted papers to share their data forms through this mechanism. Science is a team sport and the injury prevention community has a long track record of successful collaborations. We are highly supportive of this new opportunity.


Language: en

NEW SEARCH


All SafetyLit records are available for automatic download to Zotero & Mendeley
Print