Age range deemed acceptable and youngest and oldest age deemed acceptable by own age. Graphs display means and 99.5% confidence intervals

The following results for lesbian and bisexual women are based on exploratory analyses (i.e., we did not specify any hypotheses). They should therefore be seen as preliminary and be used to inform future research focusing on nonheterosexual individuals. Descriptive information as well as comparisons between groups based on sexual orientation can be found in Table S3. In addition, all results for our exploratory analyses can be found in detail in the ESM (Tables S4–S9 and Figs. S5–S14).

The subsample of women identifying as lesbian likely was too small (n ranging from 331 to 456) to determine whether age played a substantial role in shaping preference attributes (Table S4 and Fig. S5). No clear pattern was apparent for the link between age and parenting intention (Table S5, Figs. S6 and S7), yet age range deemed acceptable increased with age (Table S6, Figs. S8 and S9). As in the heterosexual sample, this was due to older women lowering the youngest age deemed acceptable in an ideal partner.

In women identifying as bisexual (n ranging from 2,239 to 3009), age was not linked to desired level of attractiveness, nor was it linked to importance and desired level of financial security and successfulness. For all other preference attributes, the subsample of bisexual women was too small to reach a definitive conclusion (Table S7 and Fig. S10). While the pattern for the relationship between age and parenting intention was mixed for bisexual women (Table S8, Figs. S11 and S12), for age range, the same pattern as for heterosexual and lesbian women was apparent: older women increased their range by being more accepting toward younger potential partners (Table S9, Figs. S13 and S14).

Discussion

The main goal of this study was to conceptually replicate and extend the study by Schwarz and Hassebrauck (2012) investigating age effects on partner preferences. Analyses were based on a large international sample of heterosexual single women, and extensive robustness checks were provided. Decisions on hypotheses for age suositellaan luet effects on partner preference attributes, parenting intention, and age range deemed acceptable are summarized in Table 9. We found no evidence for substantial age effects on any of the partner preference attributes except for a positive association between age and confidence-assertiveness. Considering the relationship between age and parenting intention, analyses suggested a linear and a quadratic relationship. In addition, higher age was linked to a broader age range deemed acceptable, with an increased acceptance for younger partners but constant levels of acceptance for older partners. This latter pattern was also apparent in exploratory analyses of lesbian and bisexual women.

Partner Preference Attributes

Supporting our hypotheses H1, H2, H3, and H5, we found evidence for no or only negligible relationships between age and almost all preference attributes, measured as both importance ratings and preferred levels of attributes. Contrary to our hypotheses H4a and H4b, we found support for a positive relationship between age and confidence and assertiveness.

These results conceptually replicate previous findings by Schwarz and Hassebrauck (2012) with one exception. Although we found a positive relationship between age and the preference for confidence-assertiveness (potentially positive for importance of confidence-assertiveness and clearly positive for the level of confidence-assertiveness), the original study found no relationship between age and conceptually similar constructs such as being dominant, cultivated, or humorous. According to Abele (2003), confidence and assertiveness are stereotypically male characteristics. One possibility is that women’s preferences for these characteristics indeed increase with age. However, given the cross-sectional nature of our data, a cohort effect with the youngest women in the sample simply being less attracted to such stereotypically male characteristics seems equally plausible. Future research utilizing longitudinal data is needed to separate age effects from cohort effects.