Paper
Sunday, November 4, 2007

334
This presentation is part of : Family Caregiving Techniques
Application of the Multi-Trait Multi-Method Technique to Multiple-Respondent Family Level Data
Charles Larew, PhD, RN, Health Promotion & Development, University of Pittsburgh, Pittsburgh, PA, USA
Learning Objective #1: describe the conceptual basis for the multi-trait multi-method technique for examining convergent and divergent validity.
Learning Objective #2: identify evidence of convergent and divergent validity for measures of family level traits collected from multiple respondents.
The multi-trait multi-method technique guided the examination of convergent and divergent validity when measuring family level traits with data collected from multiple respondents.   A cross-sectional design was utilized.   Family members of traumatic brain injury (TBI) survivors quantified pre-morbid levels of family assistance across four domains of caregiving: personal care tasks, household tasks, financial support, and assisting with health care decisions.  The Family Inventory of Life Events (FILE) assessed family demands.  Family resiliency factors were measured by the Family Inventory of Resources for Management (FIRM), Family Hardiness Index (FHI), the Family Crisis Oriented Personal Evaluation Scales (F-COPES), and the Family Problem Solving Communication Index (FPSCI).  

The sample consisted of 29 caregiver pairs.  The majority of caregivers were Caucasian (75.9%), female (69.0 %); parents or spouses (55.2 %); with less than an associate degree (53.4 %).  Subject age ranged from 22 to 73 years (mean=45.5, SD=12.0).

Strong correlations were seen regarding pre-morbid assistance with personal care activities (r=.86, p<.01), financial support (r=.89, p<.01), and assistance with health care decisions (r=.76, p<.01).  Perceptions regarding assistance with household tasks were weakly correlated (r=.38, p<.05).   Moderately strong correlations (r=.50, p<.01) were seen in FILE scores.  FIRM scores were strongly correlated (r=.73, p<.01).  Moderately strong correlations were seen with FPSCI scores (r=.56, p<.01) while weaker relationships were seen between FHI scores (r=.39, p<.05).  High correlations (r >.75) were found among FIRM, FPSCI, FHI, and F-COPES scores.  Limitations include a convenience sample and small sample size.  Low divergent validity for measures of family resiliency factors suggests a lack of conceptual distinction.  The data suggest that a single family representative may provide valid information regarding family demands and resources.  These findings may have implications for sampling and instrument selection although additional studies are needed to determine whether similar patterns of convergent/divergent validity are replicated in other caregiver samples.

See more of Family Caregiving Techniques
See more of The 39th Biennial Convention (November 3 -- 7, 2007)