Roberta Ann Pointer Smith

Expectancies about Control Over Health Relationship to Desire for Control of Health Care

Data from four studies of adults are used to address the relationship between expectancies for control of ones health (an outcome) and preference for control of health care (a process). The former is operationalized by the use of the Multidimensional Health Locus of Control (MHLC) Scales; the latter by the B and I subscales of the Krantz Health Opinion Survey (KHOS). Descriptive data from the four studies are presented. Significant correlations exist between the B subscale and the PHLC and IHLC scales. Factor analysis of datafrom three of the studies reveals that, along with selected items from other scales, all of the items from both the PHLC scale and the B subscale load on the first factor. Both theoretical and methodological reasons for these findings are discussed.

A Comparison of Terminally Ill Persons at Various Time Periods to Death

A cross-sectional comparison was made of the responses of 1,110 terminally ill persons with cancer at varying time periods prior to death. In addition, a longitudinal analysis on a subgroup of this sample was conducted for self-reported mood changes and social interaction changes as death approached. Two theories, Disengagement and Objective Self-Awareness, which suggest that people nearing death would increase in their desire to separate from others, were not supported. In general, there were few differences in the responses of persons further from, or closer to, death. The mention of religion as a source of support was less common among persons closer to death. Finally, awareness of ones demise had an effect only on ones responses regarding the role of other people in the life of the terminally ill person.

Commentary by Smith

When reading an article in a well-respected journal such as Western Journal of Nursing Research, one assumes that the research exemplifies the best of the scientific method and that there are few, if any, flaws. This important article has many strengths, including the use of a categorical model for analyzing and interpreting their data. The study, however, also contains several, perhaps unavoidable, flaws. So why was it published? This article is an outstanding example of research done well within the constraints of the research context. I was left with many questions, however, that I thought the researchers should have answered. When considering the microenvironment, why did they not examine how the developmental level, or at least age of the participants, affected their responses? Because more than 90% were less than 13 years old (those who gave their age at the time they began farmwork), how did the length of time working affect responses? How did those who had experience differ from those with little or no experience? How did the responses of those who had received EPA (Environmental Protection Agency) required training differ from those who had not? Three regions in Oregon were represented, and mean ages between groups differed. Did region and age affect responses? Selection bias is possible given the nonrandom recruitment procedures. Such bias could lead to either an overor underestimation of the problems and affect perceptions associated with pesticide exposure. Presuming that the three groups are independent of each other, looking for differences or similarities between groups, and between those trained and untrained about pesticides’ risks and precautions, could provide insight about the effects of some selection factors. The researchers had the information to answer these questions, however, they did not do so. Is that because this is a qualitative study with a very different intent than to satisfy my quantitative interests? Did the researchers feel constrained by the limited power resulting from the small sample size? Or is it because the study’s findings were so powerful that getting it into the public domain was more important than quantitative explorations? I hope that these issues will be addressed in a follow-up article. I do not have answers to my questions. I do believe, however, that the significance of the research question, the generally sound approach to collecting and analyzing data that are difficult to obtain, and the importance of the findings are sound reasons to publish this study. No study is perfect. This study raises many questions and provides