By the year 2000, more than 19 million people will be 65 years of age or older, and because the sex ratio for the oldest old has become increasingly unbalanced in the last 50 years (Bould, 1989), the vast majority of this population will be women. One of every 14 people in the United States will be a woman over the age of 65 (Uhlenberg, 1979). Robinson (1986) states that these are all reasons for studying older women, but "this attention has not been forthcoming." Even in nursing, there is a lack of research on elderly women. Matteson and McConnell (1988) note that "nurses should be aware of the inadequacy of the gerontological research base with regard to women."
A literature search located no articles that specifically focused on issues in conducting research with older women residing in the community. Much has been written about older persons living in nursing homes; in fact, according to Werley (1983), nursing home populations have been the primary focus of gerontological and nursing research. The literature revealed some of the problems researchers face when they select a population of older people to study (Bell, 1987; Duffy, 1989; Engle, 1986; Gueidner, 1989; Rowe, 1988; Zimmer, 1985). Some of these problems, for example, physical changes associated with aging, also apply to research with community-based elderly women; however, in other ways these women constitute a unique population.
This article will identify problems and provide helpful strategies for research with older women who live in the community. The article is divided into five parts: recruitment, site and sample, informed consent, measurement instruments, and strategies. The strategies are based on a literature review, research studies (Burnside, 1989; 1990a; 1990b; 1990c), and a project by two master's students (Preski, 1986). All the nursing interventions were designed to study the effectiveness of reminiscence groups with older women living in the community.
Although there are several important considerations that influence the conduct and evaluation of clinical gerontological studies, the most important concerns subject selection (Rowe, 1988). To reach a representative sample of older people is one of the "most difficult and challenging problems in gerontological research" (Gueldner, 1989). It is not difficult for a researcher to decide on potential subjects of elderly women and possible research sites; however, it is extremely difficult to obtain older women as subjects. Bell, May, and Stewart (1987) offered suggestions for recruitment: announcements at senior citizen meetings, newspaper notices, and notices on bulletin boards at sites where senior citizens reside or gather. It was our experience that sign-up sheets and bulletin boards are not always read by the persons who one would like to recruit for the research.
Researchers often face the problem of a too small sample size in their studies. The authors found that one of the most important factors in obtaining adequate sample size was the cooperation, attitude, and understanding of the managers of retirement homes and apartment complexes. Individuals prepared at the master's level in gerontological social work or nursing were effective in supporting the researchers and assisting in overcoming hurdles. They understood both the group process and the research process. They also understood the demands placed on researchers by academic requirements and Human Subjects' Rights Committees. Although it might not be possible to obtain the credentials of the managers of housing units, it is worth the time and effort to learn about their backgrounds because of their impact on the acquisition of subjects. The Figure provides a guide for the necessary groundwork for recruiting subjects who live in retirement homes or apartment complexes.
There is no substitute for personal contact by the researcher, not only with the potential subjects but also with the managers. It is extremely important for the researcher to establish a rapport with the gatekeepers. Based on our experiences, the more positive, understanding, and enthusiastic the managers of the sites were, the more successful the recruiting process.
SITE AND SAMPLE
The testing place or location of the study requires careful consideration. A meeting place with a private area that is free of distractions is a high priority. At one residential facility for the elderly, Preski and Young (1986) held group meetings in the chapel; this was the only area in the facility that was private and large enough to accommodate a group. The meeting place also has to be centrally located and accessible to members in wheelchairs and those who use walkers. One cannot assume that because the subjects are communitybased there will be no mobility problems. The distance the subject has to travel to a meeting place becomes a long way when one is 93 and moving along with a walker.
Socioeconomic and health status must be considered in research with community-based elderly women. Unless older women are selected from subsidized housing, one may find residents who are of a higher socioeconomic status than the average elderly woman (Bell, 1987). In addition, these subjects may also have a higher educational level, which could skew the research data (May, 1982). Another important consideration is the health of the subjects. Communitybased women may not be as healthy as the researcher assumes. It is possible that selecting subjects who live in the community may result in obtaining subjects with fewer disease states; however, data from several studies indicated that subjects had problems with vision, hearing, and mentation, in addition to psychological impairments (Burnside, 1990b, 1990c; Zimmer, 1985).
It is important to use a large sample in research with elderly subjects because of the serious problem of attrition; in the interim between preand post-testing, subject attrition may be caused by the illness, death, or relocation of subjects. A researcher may also mistakenly assume that older women have considerable time on their hands. However, it has been our experience that they guard their schedules with great zeal. Four commitments were prevalent in their personal schedules: doctor or therapy appointments; hair appointments; television programs, especially the afternoon soap operas; and providing care for ailing husbands and babysitting grandchildren. Working around these various commitments proved to be a challenging feat.
One group of people has not been mentioned in the literature we surveyed: the informal leaders in the groups of older women. If the informal leaders in an elderly community are recruited for the research project first, then it may become a status symbol to be in the group, and recruitment of the other subjects will be easier. In one reminiscence study, two groups of subjects were lost because of the strong influence of the informal leaders (Burnside 1990a). In one instance, the informal leader of the group was absent at the introductory meeting to explain the research. At the second meeting, when the facilitator was ready to begin the group, the informal leader was present and flatly announced that there would be no group; the women who had come for the group that day acquiesced. In the second instance, the potential subjects were undecided about whether they wanted to participate. When the researcher explained the need for demographic data, one woman became indignant and said she would not participate; the other women who had assembled then said they would not participate either.
Guidelines for Initial Contact With Residential Managers
Obtaining informed consent is a barrier to conducting research with elders, and there is little data reported in the literature that will help the investigator avoid problems and facilitate the recruitment process (Duffy, 1989). The older person's reluctance to sign papers has been discussed by other researchers (Duffy, 1989; Gueidner, 1989); Kaye, Lawton, and Kaye (1990) found that older subjects were more reluctant to sign consent forms than younger subjects. At best, elderly subjects need more time than their younger counterparts to read the consent forms and ask questions. Elderly subjects may sign an informed consent if they are in an institution because they perceive it as the desire of the physician, the nurse, or the institution (Davis, 1981). But in the community, they will not have the same perspective as a person in an institution who is dependent on others for health care.
Because the subjects in the studies discussed here were quite independent, few had been cautioned by relatives not to sign anything, as was reported in two articles (Duffy, 1989; Gueldner, 1989). The subjects were, however, unsure of what was expected of them. It was important to build a trusting relationship with the potential subjects; they needed to understand what was being done and what was expected of them.
We found it was necessary to explain the purpose of the research study several times: what their role would be, the length of study, and who would be involved. In addition, many of the subjects had physical limitations that made them reluctant to participate in a group; they had to be reassured that conditions would be safe, comfortable, and not too demanding or tiring.
Various psychological problems, for example depression and personality disorders, made some members reluctant to join the group. Other real or imagined limitations, for example education or socioeconomic status, may cause potential members to worry about being compared with others. It was deemed important, therefore, to read the entire informed consent to each group, no matter how well educated the group appeared to be.
Lastly, and perhaps the most important consideration, is that the women need to know how this research will benefit them (or women in general). The researcher must be prepared to explain the purpose of the research project many times and to restate it using different words. The bottom line is, "Why should I do this?" Researchers need to be very clear on this issue before approaching potential subjects.
Another problem in research with older women is finding the most appropriate measurement instruments and making sure that the elderly subject is able to use the instrument. Instruments may not be valid or reliable measures because the reliability and validity of the tools were established on younger populations (Burnside, 1992 Gueldner, 1989). From a different perspective, although the instrument may be appropriate for this population, a physical or educational deficit may prevent the subject from using the tool. Visual acuity and reading level are important considerations when using questionnaires with an elderly population. Subjects may be unable to see the questionnaire, or they may claim they cannot see the questionnaire because they cannot read (Burnside, 1992). In either case, their statement is taken at face value and the questionnaire is read aloud.
It is very important to do a pilot study on any research with older women (Prescott, 1989). Two pilot studies completed prior to larger studies indicated that the pre- and posttesting would be problematic because of the terminology used in the questionnaires, the inability of the subjects to fill in answers, the stress caused by the testing process, and fatigue (Burnside 1989; 1990b). In one study, Burnside (1990a) pretested the instruments on older persons of different decades; a mean for the time it took to complete the instruments was used with subsequent subjects. The subjects had a tendency to look at the questionnaires of those seated beside them, and they needed to be assured that whatever they put down was a right answer for them and not to worry about any wrong answers. The one-to-one testing method is ideal, but it is costly and timeconsuming. In addition, data may be influenced by the setting in which it is gathered. Gueidner and Hanner (1989) caution that "a disadvantage is the tendency of subjects to become involved in reminiscent conversation." Individuals who are isolated may find the interviewer an ideal sounding board. We found that subjects had a strong tendency to explain why they chose to answer an item in a certain way - a practice that was often reminiscent in nature. Unless all the data is gathered in a oneto-one setting, it must be cautioned that this "special" treatment could affect the individuals' scores.
In addition to attempting to standardize the testing process and avoiding subject fatigue, specific strategies that facilitated the testing process are helpful. The site of the research is important to the subjects. Can they easily get to the meeting place? It is important to allow enough time to phone subjects who may have forgotten to come to the testing place. The researcher must also allow enough time to allow group members with mobility problems to make their way to the testing site at an unhurried pace. Income is often a very sensitive subject with older adults; it may be possible to omit this item from the demographic data if all of the subjects live in the same type of community facility. The word "research" may intimidate some subjects; therefore, the word "project" may be used when describing the study. Specific strategies for research with community-based elderly women include:
* Be mindful of the patient's energy level and the overall length of the session when selecting the instruments.
* Validate instruments on older adults before using (Gueidner, 1989).
* Provide a quiet, relaxed atmosphere; minimize interruptions (Gift, 1987).
* Be sure the subjects remember their hearing aids and glasses and that the hearing aids are functional (Gueidner, 1989).
* Use large type on all questionnaires; use black print on white paper; select paper with a matte rather than a glossy finish (Gift, 1987).
* Place instructions and test items on one side of the page only.
* Use brightly colored pages as a cover and between the various tests in the packet to help subjects locate their place more easily during testing.
* Be prepared to respond to questions and to repeat instructions.
* Offer special assistance to anyone who needs help.
* Check every item on every questionnaire carefully at pre- and post-testing to catch omissions; double checked answers for an item, or answers placed in the wrong place on the page.
This last word of caution cannot be emphasized too strongly. As difficult as it is to obtain subjects, the researcher cannot afford to lose data because careful double checking was not done on each subject's answers. As a final note, treat the subjects with respect and tact, especially if corrections or additions are needed on thenanswer sheets.
This article described problems and strategies for researchers working with older women who live in the community. There is a lack of information in the literature about research strategies with older women living in the community. It is hoped that this article will encourage readers to study this understudied and steadily growing population. Their needs cannot be adequately addressed without research. Gerontological nurse researchers must design studies that accommodate the physical and psychological needs of older adults, and especially older women (Gueldner, 1989).
- Bell, I, May, E, Stewart, R. Clinical research in the elderly: Ethical and methodological considerations. Drug Intelligence and Clinical Pharmacy 1987; 21:1002-1007.
- Bould, S., Sanborn, B., Reif, L. Eighty-fixte plus: The oldest old. Belmont, CA: Wadsworth, 1989, pp 40-41.
- Burnside, I. The effect of reminiscence group therapy on depression and affect in older women: A pilot study. The University of Texas at Austin, 1989. Unpublished manuscript.
- Burnside, I. The effect of reminiscence groups on fatigue, affect, and life satisfaction in older women. Unpublished doctoral dissertation. The University of Texas at Austin, 1990a.
- Burnside, I. The effect of reminiscence groups on fatigue, affect, and life satisfaction in older Australian women: A pilot study. Phillip Institute of Technology, 1990b. Unpublished manuscript.
- Burnside, I., Guy, J., Koch, S., Giliotti, E The effect of reminiscence groups on fatigue, affect, and life satisfaction in older Australian women: A replication study. Phillip Institute of Technology, 1990c. Unpublished manuscript.
- Burnside, I., Preski, S., Gallman, L. Research with older adults: The Importance of the instruments. Submitted to Applied Nursing Research 1992.
- Davis, A. Ethical considerations in gerontological nursing research. Geriatr Nurs 1981; 2:269-272.
- Duffy, L., Wyble, S., Wilson, B., Mues, S. Obtaining geriatric patient consent. Journal of Gerontological Nursing 1989; 15(1):21-24.
- Engle, V. Bridging the research gap between acute and long-term care of older adults. Image 1986; 18:148-150.
- Gift, A., Palmer, M. Planning clinical nursing research with a geriatric population. Clinical Nurse Specialist 1987;1(2):57-60.
- Gueldner, S., Hanner, M.B. Methodological issues related to gerontological nursing research. Nurs Res 1989; 38:183-184.
- Kaye, J., Lawton, P., Kaye, D. Attitudes of elderly people about clinical research on aging. Gerontologist 1990; 30:100-105.
- Marteson, M. A., McConnell, E. Gerontological nursing: Concepts and practice. Philadelphia: WB Saunders, 1988, pp 481-527.
- May, E, Stewart, R., Hale, W., Marks, R. Prescribed and non-prescribed drug use in an ambulatory elderly population. South Med J 1982; 75:522-528.
- Prescott, P., Soeken, K. The potential uses of pilot work. Nurs Res 1989; 38:60-62.
- Preski, S., Young, A. The use of touch and structured reminiscence in group therapy with elderly clients. The University of Texas at Austin, 1986. Unpublished manuscript.
- Robinson, K. Older women: A literature review. J Adv Nurs 1986; 11:153-160.
- Rowe, J.W., Besdine, R. W. Geriatric medicine, 2nd ed. Boston: Little Brown & Co, 1988, pp 513-517.
- Uhlenberg, P. Older women: The growing challenge to design constructive roles. Gerontologist 1979; 19:232-241.
- Werley, H.H., Fitzpatrick, J.J. Annual review of nursing research: I. New York Springer Publishing Co, 1983, pp 75-99.
- Zimmer, A., Calkins, E., Hadley, E., Ostfeld, A., Kaye, J., Kaye, D. Conducting clinical research in geriatric populations. Ann Intern Med 1985; 103:276-283.