When fertile, women seek status via prestige but not dominance.

All social groups feature hierarchical differences, such that some people are ranked high on a valued dimension and others are ranked low. These hierarchical differences are present in all human societies, facilitating group cohesion, conflict resolution, the allocation of critical resources, and opportunities for mating ( 13 ). The ranking or relative social standing of a person on a valued dimension affects their prominence and influence within a group, thus comprising their social status. High-status individuals are more respected and admired than low-status individuals; they have greater influence over group decisions and are deferred to by others ( 4 ). The dual-pathway model of status attainment indicates that human status is allocated along two dimensions: dominance and prestige ( 14 – 16 ). Dominance involves attaining high rank through power, intimidation, coercion, and fear ( 17 ). The concept is analogous to dominance among nonhuman primates, where repeated patterns of agonistic encounters stabilize rank relations in many groups ( 18 ). Attaining status via dominance involves the use of force and threat to ensure others submit to one’s will; subordinates are obliged to defer to dominant individuals or else risk being harmed. In contrast, attaining status via prestige involves increasing one’s instrumental social value by demonstrating competence in a valued domain ( 19 , 20 ). As a result of enhanced skills and knowledge, prestigious individuals are preferred social partners, voluntarily deferred to by others, and preferentially learned from ( 21 ). Although many animal species show evidence of model-based biases for social learning ( 22 ), prestige is considered a phylogenetically recent status-seeking strategy that may be unique to humans ( 13 ). Though both sexes utilize dominance- and prestige-based strategies to gain higher rank ( 14 , 23 – 25 ), men seek status via dominance more frequently than women, and women seek status via prestige more frequently than men ( 26 , 27 ). Men are also more likely to engage in conspicuous public contests for status, whereas women’s competitive tactics are more subtle, safe, and solitary ( 28 ). Sex differences such as these have contributed to relatively fewer investigations of human status seeking among women than men. As a result, we know comparatively less about what women do when they strive for status, how high status manifests in them, and what high status enables them to accomplish. Despite these gaps, it is evident that both men and women equally desire high social regard and benefit from holding prestigious positions in a social hierarchy ( 4 , 5 ). Consistent with an evolved basis for status pursuit in both men and women, high status can also translate into higher reproductive success for both sexes ( 29 – 38 ), albeit effects for males are evident in more species, larger, and more robust ( 39 ).

To shed light on the biological correlates of status seeking among women, I implemented a worldwide, online citizen science project that tracked covariance between menstrual cycle characteristics and status seeking, dominance contest outcomes, and self-esteem. I measured women’s motivation for prestige and for dominance, thus providing a crisper picture of how fertility affects dual pathways for status attainment. Participants were from 14 countries across Europe and North America, Australia, Latin America, East and Southeast Asia, and Central and South Asia. The design employed here used best practices that rectified several common methodological challenges identified in the menstrual cycle literature: Namely, sampling multiple time points across the menstrual cycle and modeling fertility continuously provide a clearer picture of underlying patterns across the cycle compared with sampling at only fertile and nonfertile phases ( 59 , 60 ). The decreased reliability of fertile-phase identification was also compensated for by the increase in sample size ( 55 ). All measures, materials, data, and analyses relevant to this paper are available on the Open Science Framework (OSF) ( 83 ). Women began participating at any point during their menstrual cycle. After a prescreening survey and a baseline survey, participants completed a brief online survey each day for at least 28 d, and then completed a follow-up survey a week later (median number of entries 22, M = 22.44, SD = 9.20, range 1 to 49). Data were collected across 339 menstrual cycles: 68% of women opted to provide data for one cycle, 28% opted to provide data for two cycles, and the remainder provided data for three cycles. I measured the motivation for dominance and prestige using items derived from the Dominance and Prestige Scale ( 15 ), a well-validated scale with good construct and internal validity that measures the dual pathways to status seeking. For dominance, participants rated how much they enjoyed having control over other people, how willing they were to use aggressive tactics to get their own way, and whether they tried to control others rather than permit others to control them. For prestige, people indicated how important it was that they were respected by others, admired for what they had done, were successful, were recognized for their achievements, were able to show their abilities, and that people did what they said. * To measure dominance contest wins then losses, participants answered four questions about incidental occurrences during the day being diarized: “If I had a conflict with a man/woman, afterwards I came away feeling like I (he/she) got the upper hand” and “If I thought a man/woman was out of line, I (he/she) put him/her (me) in his/her (my) place.” If participants did not experience a contest that day, they were able to leave these questions blank. Self-esteem was assessed via the State Self-Esteem Scale ( 61 ), which measures self-perceived competence (performance self-esteem), an individual’s personal evaluations about relationships with others in their lives (social self-esteem), and a person’s appearance, attractiveness, and body image (appearance self-esteem). To reduce experimenter degrees of freedom, all decisions concerning the calculation of menstrual cycle variables and the analytic approach outlined here were preregistered for another investigation on the OSF ( https://osf.io/zw8qx ). I used the backward-counting method to estimate conception probability, assigning each backward-counted day a specific conception probability based on the estimates from Stirnemann et al. ( 62 ), a method advocated by Gangestad et al. ( 59 ). Backward-counting methods are more reliable indicators of cycle phase than forward-counting methods, as the length of the follicular phase is more variable than the length of the luteal phase. Many women experience negative psychological and behavioral symptoms around menstruation, including low mood, shame, and physical pain ( 63 ). These symptoms are compounded by menstrual taboos and a lack of affordable sanitary products in many environments, both of which can also increase social isolation. Poor mood, body shame, physical symptoms, and social isolation are all likely to affect social interactions, including those related to status seeking and self-esteem. To ensure that any status effects investigated here were not due to menstruation, participants indicated whether they were currently experiencing menstrual bleeding each day. Following best-practice protocols ( 55 ), I controlled for menstruation as a fixed and random effect in all analyses, thus accounting for any potential effects of menstruation on status seeking.

Bayesian mixed models controlling for menstruation indicated that the motivation to acquire status via prestige, dominance contest losses, appearance-based self-esteem, and performance-based self-esteem all positively covaried with conception probability ( Tables 1 and 2 ). There was no effect of fertility on dominance contest wins, acquiring status via dominance, or social self-esteem. When fertile, women sought more status via prestige, experienced more dominance contest losses, and experienced higher self-esteem in their appearance and performance. Menstruation estimates were not significant for any outcome except for a negative effect on prestige that was roughly half the size of the fertility effect. These estimates are in SI Appendix , Table S1 . Main effect of conception probability on status seeking and dominance contest outcomes To control for between-individual variance and the potential effects of menstrual symptoms on the outcome variables, all models included a random and fixed slope for menstruation and a random intercept for subject, in addition to a random slope for conception probability. * Significant effect. CI L = Lower credible interval. CI U = Upper credible interval. Main effect of conception probability on self-esteem To control for between-individual variance and the potential effects of menstrual symptoms on the outcome variables, all models included a random and fixed slope for menstruation and a random intercept for subject, in addition to a random slope for conception probability. * Significant effect. CI L = Lower credible interval. CI U = Upper credible interval. To visualize outcome variation continuously—without imposing discrete phases such as menstruation and the fertile window upon the data—I fit Bayesian mixed models with a Gaussian family and cyclic cubic splines over backward-counted cycle days. As shown in Figs. 1 and 2 , all four of the variables yielding a significant fertility effect visually peaked just prior to the estimated day of ovulation. Status seeking via prestige, status seeking via dominance, dominance contest wins, and dominance contest losses plotted across the menstrual cycle. The dashed line indicates the estimated day of ovulation and red indicates a significant fertility effect for this variable. Each plot is the result of 100 random samples taken from the posterior of a Bayesian mixed model with a Gaussian family and cyclic cubic splines over backward-counted cycle days. They visualize the conditional means for the continuous splines, reflecting average patterns in the data after including uncertainty from the person-level clustering. For slight regularization, half-normal priors were set with an SD of 1 on the random intercepts. Self-esteem plotted across the menstrual cycle. The dashed line indicates the estimated day of ovulation and red indicates a significant fertility effect for this variable. Each plot is the result of 100 random samples taken from the posterior of a Bayesian mixed model with a Gaussian family and cyclic cubic splines over backward-counted cycle days. They visualize the conditional means for the continuous splines, reflecting average patterns in the data after including uncertainty from the person-level clustering. For slight regularization, half-normal priors were set with an SD of 1 on the random intercepts. To test the robustness of the significant effects, I examined whether they were moderated by age, cycle length, or fertility awareness. Dominance varies across the lifespan and menstrual cycle effects vary by age ( 64 , 65 ). I tested for moderation by cycle length, as very long or very short cycles are less likely to be ovulatory ( 66 ), and tested for moderation by fertility awareness to reduce expectancy biases. I entered predictors as a main effect and moderator of conception probability and menstruation (three models, one each for age, cycle length, and fertility awareness). No significant effects reported here were moderated by these variables (all credible intervals contained zero). Questions concerning dominance contest outcomes referred to bouts with male and female competitors, yet analyses collapsed across competitor sex. Because past work has indicated contest behavior and contest outcomes are affected by the sex of the competitor ( 6 ), I tested whether competitor sex moderated the effect of fertility on contest outcomes. I added competitor sex as a fixed main effect and moderator of the fertility effect, but no moderation by competitor sex was detected (all credible intervals contained zero). I then examined whether the likelihood of entering into a contest—rather than the outcome of the contest itself—depended on conception probability. I coerced the contest outcome variables into a binary variable, such that any score on the win or loss variables reflected the occurrence of a dominance contest that day and missing data reflected the absence of a dominance contest that day. Using a Bayesian mixed model with a Bernoulli family and logit link, I regressed the binary contest occurrence variable onto conception probability and menstruation, including maximal random effects and a random intercept for subject. Conception probability did not covary with the likelihood of entering into a dominance contest (credible interval contained zero). I then examined whether the pattern of effects reported here was evident among a sample of hormonal birth control users who had also submitted data for studies in the daily cycle diary. I followed the same analytic procedure as outlined here but for this separate sample. Conception probability did not predict any outcome variable for hormonal birth control users (all credible intervals contained zero; details are in SI Appendix ). I also examined whether outcomes were moderated by socioeconomic status, relationship status, or geographic region of residence. I entered these covariates as a main effect and moderator of conception probability and menstruation. No significant effects reported here were moderated by these variables.

Investigations of biological correlates of status seeking often focus on variation in testosterone and cortisol. The most prominent theory in this area is the dual-hormone hypothesis, which posits that testosterone positively predicts status seeking when cortisol is low ( 40 ). Evidence for the dual-hormone hypothesis has been mostly supportive, but there are unresolved questions regarding the size of the effect and its generalizability ( 41 ). Stronger support for the hypothesis is found for men compared with women ( 42 ), a finding which may reflect sex differences in the reliability of testosterone immunoassays or that other hormones affect status seeking in addition to testosterone and cortisol ( 43 ). Estradiol—an estrogen that is the primary reproductive hormone in females—is a likely candidate for one of these other hormones. Though the mechanism of action is likely different from that of testosterone, estradiol increases dominance behavior in several mammals and some evidence suggests it is associated with the behavioral activation system in humans, a system concerning reward sensitivity, appetitive motivation, and sensation seeking ( 44 ). Though no studies directly investigate the relationship between estradiol and status seeking in women, convergent findings suggest one. Women’s estradiol positively covaries with their assertiveness, self-efficacy, and power motivation ( 45 – 49 ) (but cf ref. 50 ). Instead of measuring estradiol directly, some investigations have examined covariation between conception likelihood and status-relevant outcomes. Estradiol peaks during the periovulatory period when women are most likely to conceive; in a standard 28-d menstrual cycle, this is the 5 d preceding ovulation and the day of ovulation itself ( 51 ). These studies have demonstrated that fertile women prefer luxury consumer goods that confer high status, are especially interested in maximizing their economic capital and winning economic games, and are quick to derogate female rivals ( 6 – 10 ) (cf refs. 52 – 55 ). Scholars propose that periovulatory psychological and behavioral shifts facilitate adaptive goals by providing advantages in mate competition ( 56 , 57 ). High status can also facilitate adaptive goals by making females more desirable allies, ensuring they have better access to critical resources and social networks, and improving their ability to advocate for their own interests ( 29 – 33 , 35 – 37 ). Based on this prior evidence, periovulatory women may experience heightened status seeking. If fertility is a biological correlate of status seeking, then the periovulatory window likely confers tangible benefits to naturally cycling women. Understanding these connections could benefit all women experiencing menstrual cycles, especially the 100 million women using hormonal contraceptives (HCs) every day who do not experience cyclic fertility ( 58 ). Though the presence of fertility effects in naturally cycling women does not necessarily indicate their absence in HC users, it does suggest that such effects may be missing among this population.

Fertility positively covaried with the universal psychological drive to acquire status via prestige but not the desire to acquire status via dominance. This result suggests that conceptive women are especially interested in being seen as competent social partners whom others admire and defer to but not in gaining status through coercion or fear. Women experienced more dominance contest losses when fertile but not more dominance contest wins, a finding paradoxically consistent with the motivation to obtain and sustain high prestige. State self-esteem increased when conception probability was high, but the effect was specific to the dimensions of performance and appearance and not social self-esteem, the latter of which lacked construct validity. All variables positively covarying with fertility peaked during the periovulatory phase just prior to the estimated day of ovulation.

Status seeking is a fundamental human motive and its biological correlates have important implications for sex differences and similarities in human competition and the psychology of social rank. Here, I investigated whether the motivation to obtain status via prestige and via dominance, dominance contest outcomes, and self-esteem varied across the menstrual cycle in 253 naturally cycling women from 14 countries. An analysis of 4,179 daily diary entries controlling for menstruation indicated that status seeking via prestige, dominance contest losses, appearance self-esteem, and performance self-esteem positively covaried with conception probability. These variables peaked during the periovulatory phase just prior to the estimated day of ovulation. There were no cycle-related effects for status seeking via dominance, dominance contest wins, or social self-esteem. A range of studies have demonstrated that fertility affects mating competition, fueling the desire to attract mates via self-promotion and rival derogation ( 6 – 8 , 48 , 55 – 57 , 67 – 70 ). Studies on this topic are theoretically grounded in the adaptive benefits of high competition when conception is likely, namely allowing women to seek and secure the highest-quality mates when it matters most. The results reported here extend this work by showing that fertility effects are broader than previously theorized, extending to the universal psychological drive to acquire status via prestige. This result suggests that fertility motivates behaviors designed not only for mating competition but also for gaining status and maintaining positive regard in social hierarchies. In recent years, large-scale replications using longitudinal methodology and impressive analytic techniques have cast doubt on the replicability of ovulatory shifts in a range of variables, including some aspects of mating competition ( 52 – 55 ). In contrast to previous supportive literature, these studies converge to suggest that fertility does not increase self-report rivalry toward intrasexual competitors, derogation of rivals via facial attractiveness ratings, or jealousy. How can we make sense of these mixed findings? Although the current study did not specifically examine rival derogation or self-promotion aspects of mating competition, status seeking via prestige is conceptually closer to self-promotion than it is to rival derogation. Ovulatory shifts in mating competition may thus affect self-promotion to a greater extent than rival derogation or jealousy. This interpretation is consistent with work showing that fertility covaries with competitiveness concerning developing the self, self-efficaciousness, and the desire for social power ( 46 – 49 , 71 ). Prestige is freely conferred by others based on the instrumental value an individual offers to a social group; it involves deference, admiration, and respect ( 20 ). That the desire to obtain status via prestige increased alongside fertility suggests that conceptive women are especially interested in being seen as competent, effective social partners whom others admire and defer to. Such a finding is consistent with women’s overall intrasexual competition strategy, which involves promoting the self, actively defending of one’s own reputation, and inflicting damage upon the reputation of competitors ( 72 ). Holding high esteem in the eyes of others better equips women to enact this competitive strategy and achieve social capital. In contrast with fertility effects on prestige, fertility did not covary with status seeking via dominance or winning dominance contests. The effects of fertility on status seeking appear limited to affecting one’s environment through nondominant means; they do not extend to gaining status through coercion, fear, or intimidation. Of the two pathways to status, prestige tends to confers higher social rank than dominance, especially for women ( 27 ). Individuals who gain status via dominance tend to be disliked ( 14 ) and dominant behavior by women commonly evokes backlash by others ( 73 ). That fertility does not covary with dominance suggests that fertility does not blindly motivate a grab for power at any cost, nor does it motivate behaviors which may elevate current rank but erode one’s positive regard in the eyes of others. These findings point toward fertile women avoiding fear-based status-seeking strategies, preferring to attain status via social influence and admiration. Unlike null fertility effects on dominance contest wins, contest losses peaked during the periovulatory phase. One account for this effect is that other people may be more likely to engage fertile women in dominance contests or aggressive bouts. Many female mammals—including women—are especially vigilant of other females high in fertility, with some work suggesting that women preferentially and selectively aggress against fertile mating rivals ( 74 ). In contrast to these findings, however, participants here were not engaged in dominance contests more regularly when fertile and contest outcomes were not moderated by competitor sex. Thus, dominance contests during the fertile window did not manifest because women were seeking to dominate others or that others were especially seeking to dominate them. Instead, when fertile women found themselves in a dominance contest, they lost it more often than not. An alternative account for this finding is that women may intentionally lose dominance contests when fertile to maximize their ongoing high social regard. When people seek status via prestige they tend to avoid behaviors which harm prestige conferral, such as aggression and competitiveness ( 75 ). As noted earlier, people—especially women—can be perceived poorly when they gain status via dominance ( 14 , 73 ), and prestige-based status-seeking strategies tend to confer more status upon women than dominance-based ones ( 27 ). When confronted with a dominance contest, losing it may be more beneficial to women than winning it but potentially sustaining damage to their social reputation. This proposal suggests that dominance contest losses during the fertile window paradoxically reflect the same motivational drive that sees fertility increase the motivation for prestige. Investigations of the relationship between conception probability and self-esteem have yielded inconsistent results ( 55 , 76 ), as have investigations of the menstrual cycle and mate preferences ( 57 ). One methodological difference between self-esteem studies and the current work was that I measured self-esteem using an instrument that is state-level and multidimensional. Though self-esteem is often considered a global construct, there are compelling reasons to believe that global self-esteem is a synthesis of numerous components ( 77 ). Using a state-level, multidimensional measure, I find that fertility affected self-esteem in performance and appearance domains and not in the social domain. Fertility effects on self-esteem thus appear to be domain specific. The existence of periovulatory shifts in status seeking indicates that there are some benefits to status seeking that are transient. Given that high status is generally considered to provide advantages that are ongoing, one resulting question is why status seeking via prestige should be so short-lived. Here, I propose that prestige allows fertile women to be seen as competent, effective social partners whom others admire and defer to; self-promote; actively defend their reputation; and inflict damage upon the reputation of competitors. Some of these effects clearly benefit mating competition and as such should be expected to peak during the brief window of conception. Other effects confer benefits that could foreseeably improve women’s reproductive success more broadly, ensuring access to critical resources and social networks, and improving their ability to advocate for their own interests. Why might these latter effects peak during fertility? To understand these findings, it helps to remember that people spend more time admiring, watching, listening to, and deferring to the wishes of prestigious individuals ( 4 ). For women to receive this social attention during the fertile window when they feel attractive, competitive, and efficacious ( 49 , 70 , 71 ) likely improves their ability to navigate social hierarchies, compete with rivals, attract mates, and exert mate choice. To the extent that some of these status-seeking behaviors incur social costs ( 74 ), greater attractiveness and competitiveness during fertility may also mean that fertile women are better placed to incur them. Though this account is consistent with the existing literature, it would benefit from rigorous future work documenting the temporality of status benefits for fertile and nonfertile women. If status seeking via prestige enhances reproductive success as I have proposed here, some may find it surprising that relationship status and the sex of the competitor did not moderate any of the reported effects. One explanation for the null effect of relationship status is that I only tested moderation for relationship status measured at baseline. If women’s relationship status changed during the study, effects would not have been detected. Women also only reported engaging in dominance contests on ∼30% of diary entries, a finding which could simply reflect that dominance contests were infrequent, or that other contests were occurring that were not captured by the measures here. Future work examining within-person changes in relationship status and further examination of dominance contest outcomes by competitor sex would help clarify these findings. Null effects of fertility on social self-esteem were somewhat inconsistent with the other findings reported here. If fertile women are preoccupied with holding prestigious positions in social groups and increasing their social capital, one might expect that they would hold their impression management competence in high regard. The items used to measure social self-esteem here, however, were not ideally worded, even though they derived from a well-validated instrument. Instead of measuring competence in impression management as intended, the social self-esteem items appear to measure anxiety in impression management. Future work intent on dissecting the effects of fertility on self-esteem domains should incorporate measures that reflect one’s high esteem in the social world rather than one’s anxiety regarding their impression management capabilities. Some investigators have tested group differences between naturally cycling women and HC users to confirm that fertility-induced psychological variation is not evident among women who are not naturally cycling ( 55 ). Here, I examined whether fertility probability affected status seeking among a separate sample of HC users, finding that the pattern of effects reported here did not replicate in this sample. Formal comparison of this relationship is a natural extension of this work: It would be useful to quantify the size and specificity of the group difference and the influence of different types and dosages of contraception. A recent review of 46 studies indicated that HCs disrupt competitive decision making, suppress achievement-based competitive motivation, and reduce intrasexual competition under certain contexts ( 78 ). Furthering knowledge in this area is of utmost importance. Understanding more about women’s engagement in dominance contests across the menstrual cycle would also be a useful focus of future work. The dominance contests investigated here were incidental occurrences, with no data gathered on the situation eliciting the contest, the topic of the contest, or the other contestant. Documenting these criteria would provide a more nuanced picture of how women negotiate status disputes and the environments that elicit them. It is also worth noting that this study primarily consisted of self-report items and not behaviors, and thus required self-awareness of status-seeking desires. Given that women often disguise their competition with others, even from themselves ( 79 ), it is plausible that effect sizes in the current study are underestimates of the true effect. Replication of these findings with implicit and behavioral methods that do not rely on self-awareness would be a welcome inquiry to this field of work and clarify the true size of the effect.

Participants were recruited from Prolific Academic, the University of Melbourne undergraduate participation pool, and the general community worldwide for a large, ongoing citizen science project involving “the menstrual cycle and psychology, health and wellbeing” ( https://go.unimelb.edu.au/o84e ). Participants from the general community were recruited through advertisements on Facebook and Twitter and via popular science and magazine articles advertising the study and the free, personalized report provided to all users. This report compared users’ responses with the grouped average responses of naturally cycling women and HC users worldwide. Data from the citizen science project had been collected over three waves at the time of this study’s completion, with some variation among measures across waves. These measures included questions on contraceptives, affect, personality, impulsiveness, self-regulation, health, and sexual behavior. The study was approved by the University of Melbourne Ethics Committee and all participants provided written informed consent to participate. Selection criteria for eligibility in this study were fluency in English; regular menstrual cycles; menstrual cycle–length confidence exceeding the scale midpoint; reported menstrual cycle length between 22 and 35 d; aged between 18 and 45 y; premenopausal; no emergency contraception, breastfeeding, or pregnancy within the past 3 mo; no polycystic ovarian syndrome or endometriosis; and no medically diagnosed fertility or endocrine issues. These selection criteria were preregistered for another investigation on the OSF ( https://osf.io/zw8qx ). Almost a third of participants were ethnically Northern European (29.2%), 13.4% were of mixed heritage, 12.6% were from Oceania, 9.9% were Southeast Asian, 8.3% were Southern European, 6.3% were East Asian, 4.3% were South Asian, 4.0% were Eastern European, 3.2% were African, 2.8% were North American, 2.4% were Western European, 1.6% were Central, Latin, or South American, and 0.4% were West Asian. Most participants were exclusively heterosexual (68.3%), 17% were occasionally or more than occasionally heterosexual, 2.8% were bisexual, 4.4% were predominantly or exclusively homosexual, 2% were asexual, 5.2% were pansexual, and the remainder were other. Most participants reported average relative socioeconomic status (57.9%), 10.3% reported low socioeconomic status, and 31.7% reported high socioeconomic status. I implemented daily surveys with the formr survey framework ( 80 ). All data and syntax used in these analyses are available on the OSF ( 83 ). Unless specified otherwise, all items were ordinal and measured on a five-point scale (strongly disagree, disagree, neither, agree, strongly agree). One item within each measure was randomly presented each day for all measures excluding the contest outcome items, where two items were presented each day. I used Bayesian ordinal mixed models with a cumulative family and a logit-link function to model the ordinal outcomes and used a Gaussian link for the two continuous outcomes. Although it is common to analyze ordinal data with methods that assume metric responses, this process can distort effect-size estimates, inflate type I errors, invert differences between groups, and lower rates of correct detection ( 81 ). The Bayesian framework provides considerably more information about the models and their parameters than the frequentist approach and allows a superior quantification of uncertainty ( 82 ). Model estimates were considered to be significantly different from zero when their 95% Bayesian CIs did not cross zero. All models included fixed and random effects for conception probability and menstruation, as well as a random intercept for the individual.