Friendship

Most of the research on adult friendship has been conducted since the early 1970s. Early studies focused on the number of friends people had and how much time they spent with them. More recently researchers have shifted their focus to the study of other aspects of friendship structure such as what proportion of people's friends know each other, whether the friends treat each other as equals, and whether they are demographically similar to each other, to dimensions of friendship process such as feelings, thoughts and behaviors involved in a relationship, and finally to the variation in both friendship structure and process across contexts. As Adams and Allan discussed elsewhere, these changes in foci reflect the realization that friendships are complex and that they vary tremendously depending on the network, community, and society in which they are formed and maintained.

Gerontologists have examined friendship processes more closely than they have examined friendship structure. In addition to the studies of what people think about their friends, such as those on how older adults define friendship mentioned above, gerontologists have researched how older adults feel about their friends and what they do with and for them. For example, some researchers have reported that older adults feel more satisfied with their friendships when the favors they do for their friends are reciprocated, but Karen Roberto and Jean Scott found that reciprocity was less important among close friends than among casual ones. This finding has implications for the durability of friendships as people age and can no longer help others as much as they could when they were younger.

Most of the research on older friendship, however, has focused on what friends do together, such as sharing companionship, communicating with each other, and especially helping each other. Eugene Litwak noted that in contrast to family members who help older adults with tasks that require long-term commitment, friends are more likely to help older adults with shorter-term tasks or events that they share in common. For example, the friends of older adults might help them adjust to widowhood, make a decision about when to retire, and decide whether to relocate, whereas family members might nurse older adults with chronic physical problems or manage their finances. The difference in the ways which friends and family members may help older adults may have implications for the welfare of older adults without families.

The research findings on the structural features of older adult friendships are much less conclusive than those about its processes. More research has been conducted on the size of older adult friendship networks and how similar friends are to each other than how likely the friends of an older adult are to know each other. Power and status differentials between older adults and their friends have not been studied at all.

Each study of friendship reports a slightly different average number of friends for older adults. Some of the variation in findings can be attributed to differences in the contexts in which older adults live. For example, researchers commonly report that institutionalized older adults report fewer friends than those who live independently. Differences in the demographic composition of the populations studied also contribute to varied results. For example, like many other researchers, Claude Fisher and Stacey Oliker reported that older men have fewer friends than older women. This suggests that researchers who study samples of older adults in which women are overrepresented will report more friends on the average. The age composition of the sample also affects the average number of friends reported. Many early studies reported that the older adults were, the fewer friends they had. Given these findings, one would expect researchers who study populations in which the average age is high to report a smaller number of friends than those who study younger populations. It would be a mistake, however, to assume that a loss of friends with age is inevitable, universal, and linear, because other researchers such as Colleen Johnson and Lillian Troll and more recently Dorothy Jerrome and Clare Wenger, have demonstrated that some people continue to add new friends to their networks as they age.

The findings regarding the similarity of older adults and their friends and the proportion of the friends of older adults who knew each other also vary by study. It is clear, however, for older adults as well as for people of other ages, that the characteristics of a contest affect the characteristics of the networks embedded within it. For example, Pearl Dykstra and others have reported that during old age the proportion of women's friends who are women is higher than the proportion of men's friends who are men. Although this gender difference exists in all age groups, it is larger in old age, probably because women live longer and thus more of them are available to be friends. The tendency to form relationships with people who are similar to them and the relatively low proportion of men who reach old age suggests that men may be at a disadvantage in establishing new friendships and women may have difficulty developing a diverse network. As Litwak observed, a diverse friendship network is desirable because different types of friends have access to different resources and can help older adults in varied ways.

Studies of the proportion of an older adult's friends who know each other also illustrate the importance of contextual effects. Comparing results in studies of different contexts reveals that the friends of older adults in nursing homes are more likely to know one another than the friends of older adults in age-segregated housing, and that friends of older adults in age-segregated housing are more likely to know one another then the friends living in age-integrated community settings (Blieszner and Adams). The proportion of people's friends who know each other has implications for the types of help they can seek from them. Consider a situation in which an older woman expects to be bedridden for a substantial period of time. If a high proportion of her friends know one another, only one phone call may be necessary to activate a helping network. If, however, her friends do not know each other, then a whole series of phone calls may be necessary. In contrast, imagine an older man with a secret to share. If his friends all know each other, he may worry gossip will spread. If his friends do not know each other, he can be confident that his story will not be retold to anyone who matters to him.