The purpose of this study was to examine the relationship between social network diversity and three dimensions of SPA. We hypothesized that more diverse social networks are related to more gain-related SPA and less loss-related SPA. Furthermore, we assumed that this association would be moderated by chronological age. Our study found partial support for both hypotheses: Considering an interaction between chronological age and social network diversity revealed that at higher ages, older adults with more diverse social networks showed more gain-related SPA in the dimension of ongoing development than older adults with less diverse social networks. Contradicting both of our hypotheses, loss-related SPA in the dimension of social losses had a small positive (instead of negative) association with social network diversity, without any relevance of chronological age for the relationship. Also, in opposition to both hypotheses, loss-related SPA in the dimension of physical losses showed no association with social network diversity and no moderating effect of chronological age.
Several aspects of our results merit comment. Overall, similar to previous findings (Beyer et al. 2017), SPA were found to become less gain-related and more loss-related across age groups. However, our results imply that social network diversity exerts a cushioning effect on the decrease of the perception of personal development. This observation can be interpreted as partial support of the assumptions of the DIRe model, claiming that diverse social networks compensate best for age-related losses (Huxhold et al. 2022). Results of qualitative studies indicate that the loss of social roles with age-related events such as retirement challenges the sense of purpose for some of those affected (Hobbis et al. 2011; Jones et al. 2010). There is reason to assume that maintaining purpose by performing various social roles and exploiting the stimulating potential of a diverse social network contributes to perceiving ongoing development in older age. However, our results suggest that diverse social networks foster capabilities and personal development rather than a sense of social belonging. Perceiving social losses was slightly positively related to social network diversity. One possible interpretation of this finding is that, resembling the assumptions of the SST (Carstensen 1992), diverse social networks could be perceived as less emotionally rewarding than a network of very close and mostly familial ties. Likewise, initial evidence suggests that older adults are better able to cope with negativity in closer ties than in less close ties (Birditt et al. 2020). Comparably higher vulnerability to experiences in weaker relationships could lead to a higher perception of social losses among older adults in diverse social networks. The fact that social network diversity did not play a role for the perception of physical losses suggests that having a heterogeneous network exceeding close familial relationships does not contribute considerably to (the subjective assessment of) age-related changes in physical health. Past research indicates that essential support buffering against stress is provided by few close relationships (Cohen and Wills 1985) and that instrumental support to older adults is mostly provided by children (Seeman and Berkman 1988). Thus, it is likely that few close contacts such as adult children provide important assistance and fulfil basic needs in terms of physical health, already.
Our findings regarding SPA were mostly corroborated by the sensitivity analysis on age stereotypes. Remarkably, effect sizes even increased in relation to age stereotypes compared to the corresponding SPA dimensions. Concluding from these results, the extent of social network diversity is more strongly related to how the role of older adults is perceived in general (AS) than to the appraisal of the personal aging process (SPA). SPA might be rooted primarily in more profound characteristics like personality and more individual experiences such as the personal biography (Wurm et al. 2017). AS, in turn, are more likely to stem from shared experiences and may therefore be more strongly related to the personal network and its characteristics.
Limitations and directions for future research
We acknowledge several limitations of this study. One weakness of the measurement of social networks applied in the DEAS questionnaire is that it does not record actual peripheral contacts, such as occasional social interactions in everyday life. A suitable approach to this in future studies could be the Ecological Momentary Assessment (Fingerman et al. 2020). Moreover, the social network diversity score entails a simplification of network diversity. Grouping the social roles into an index cannot reflect differences between role constellations with the same index value. Furthermore, adding qualitative properties of relationships to future examinations could contribute to a more differentiated understanding of specific roles.
The data only allow conclusions about the social networks of community-dwelling older adults, due to the sampling criteria of the DEAS. Institutionalized older adults should be considered in future studies. A particular advantage of the institutional context would be to capture complete social networks of which all network members could be surveyed (Ayalon and Levkovich 2019).
Data on the SPA variables were initially biased by selective-non-response with respect to the paper-and-pencil-questionnaire. However, this limitation was corrected by the weighting procedure described in the statistical procedure.
Another limitation rests with the calculation of the mean scores of SPA domains. This approach assumes interval equality of ordinal items, which is, in fact, unknown (Kampen et al. 2000). Furthermore, critics might argue that keeping the covariates constant in all models did not account for the domain-specificity of SPA appropriately. Future studies could explore whether results differ when modelling domain-specific sets of covariates.
Lastly, while the study provides initial evidence on the association between social network diversity and SPA, further investigations are required to understand the direction of the relationship.