An Echo, a Hurrah, and Other Reflections 291an unfair advantage. Overlapping some of these points, Holt- Lunstad notes friends
can foster risky, unhealthy behaviors as well as create stress. In this general vein, King
et al. found that college students who had difficulty maintaining friendships were
more likely to report suicide attempts, drug addiction, and/ or drinking problems.
Reconciling the Pros and Cons of Friendships
Reading this volume, it is clear that friendship has both a positive and a troublesome
side, yet overall it seems to me the positive side has the upper hand. In terms of
the scholarly literature, although some (Gable & Reis, 2001) take an opposite view,
two analyses of relationship research have found that coverage of positive topics is
more common than attention to negative aspects (Duck, 1994; Hoobler, 1999).
With regard to actual friendship, previously cited data shows being with friends is
associated with positive feelings. Friendships are voluntary, so people can disengage
if they want. Friends typically do not have the obligations and responsibilities that
cohabiting and marital partners have. All in all, it is not surprising that being with
friends is generally rated as a positive time in our lives.
A key question becomes, when will friendships be positive for us either in terms
of our evaluating them positively or their leading to beneficial outcomes, and when
will they be negative? Providing a comprehensive, concrete answer to that question
is a daunting task, but I believe we have already seen some elements of the answer
and I have ideas about the form the analysis could take. Contributors to this volume
have already identified several variables that predict relationship satisfaction. The flip
side of those predictions points to when and for whom relationships are not working
so well.
Apropos of the form the analysis might take, Clark et al. classify predictors of
the success of postromantic relationship into three categories: individual, dyadic,
and social network. These categories of variables are clearly important. I would
add a fourth category: a broader array of contextual and environmental factors. In
research on the health benefits of relationships there is a lot of concern about the
form of social connectedness and the pathways via which relationships lead to out-
comes (Holt- Lunstad). There is also concern about gender and other group differ-
ences. All in all, I might frame the question about positive and negative outcome as
follows: Who, under what conditions, via which processes leads to which positive
versus negative outcomes of friendships? Essentially this boils down to various cat-
egories of variables: predictor, mediator, moderator, and outcome.
Berkman, Glass, Brissette, and Seeman (2000) offer one frequently cited model
of the association between social integration and health. They start with macro-
structural conditions that influence social networks, which in turn are a foundation
for psychosocial mechanisms (e.g., social support) that impact health via various
behavioral, psychological and physiological pathways that contribute to positive
versus negative health outcomes. Their explication nicely identifies things to con-
sider at each step in their model. For a friendship model we would need to narrow