Home Engineering Behavioral Intervention Research: Designing, Evaluating, and Implementing
Baseline Data Collection
Data collected at baseline and prior to randomization and group allocation can inform specific research questions that lead to meaningful publications. One type of publication is the evaluation of the psychometric properties of a new measure if a sufficient sample size is available. Gitlin and colleagues’ (2005) publication on the psychometric properties of the Caregiver Assessment of Function and Upset (CAFU) is an example of such an approach. The data presented was from the baseline data collection effort of the NIH Resources for Enhancing Alzheimer’s Caregiver Health (REACH I) multisite initiative involving an evaluation of six distinct caregiver interventions in diverse geographic locations (Gitlin et al., 2003). The CAFU, a 17-item measure of caregiver report of functional dependence of the person with dementia and the caregiver’s own level of upset with dependency, was developed by REACH investigators and utilized by three of six sites. Once baseline data collection was completed, an evaluation of the psychometric properties of this measure was possible. The psychometric properties were important to report as the CAFU was subsequently used as an outcome measure.
Similarly, a psychometric evaluation may be warranted if the population included in the trial represents a new group of individuals for whom validation of a previously developed scale has not occurred. For example, Roth and colleagues (2003) used the NIH REACH I baseline data set involving over 1,200 dyads (family caregivers and persons with dementia) to evaluate the psychometric properties of the Revised Memory and Problem Checklist (Teri et al., 1992). The analyses extended an understanding of its psychometric properties by confirming its factor structure with White/Caucasians, examining its properties for Hispanic/Latino and African American/Black, not previously considered, and evaluating its subscale, caregiver upset with behaviors, that had also not previously been examined.
In addition, there is always a need to develop and advance appropriate measures for intervention studies. Therefore, introducing items at baseline relevant to the study purpose and testing their psychometric properties as part of a trial can advance measurement in important ways (Gitlin, Winter, Dennis, & Hauck, 2006; Gitlin et al., 2002).
Although the main goal of a trial is to evaluate intervention benefits, there is a multitude of research questions of a correlational nature that can be posed using cross-sectional data collected at the baseline of a trial. For example, the Maximizing Independence at Home (MIND at Home) Trial examined the effects of an 18-month intervention helping 308 individuals with cognitive impairment to remain at home (Samus et al., 2014). Although main study outcomes of most interest were gathered at 18 months, important questions were also posed using cross-sectional analyses with the data collected at the baseline interview. These have included a description of unmet needs of persons with cognitive impairment (Black et al., 2013), how needs differ by cognitive impairment level (Hodgson, Black, Johnston,
Lyketsos, & Samus, 2014), and the relationship between level and type of need and caregiver burden (Hughes et al., 2014).
|< Prev||CONTENTS||Next >|