Are gender roles associated with well-being indicators? The role of femininity, fear of negative evaluation, and regret in decision-making in a spanish sample

Gender roles operate as a social schema through which people learn how they must behave and make decisions; that is, socially, what women and men are expected to do is pre-established. This gender socialization is such a deep-rooted process that people may not aware of how they have been socialized, which is thus often is difficult to analyze in people's discussions of their own decision-making. On this basis, this research examines whether endorsing gender roles could be related to women’s decisions (regret), as well as the possible consequences for their well-being. Moreover, we analyzed the content of the most important decisions that people make throughout their lives. Therefore, this research aimed to understand the meaning of gender and how its internalization seems to be related to sex differences in decision-making and well-being. In a sample of the Spanish general population (N = 203; Mage = 41.70, SD = 10.93; range from 20 to 65 years old), results showed that women (vs. men) had a greater internalization of gender roles (i.e., femininity), which was associated with higher fear of negative evaluation. This, in turn, was associated with experienced regret in decision-making, which finally seemed to lead to lower well-being. Analysis of the content of decisions showed that women’s decisions were based mainly on work and family domains, whereas men’s decisions were based on work to a greater extent.

People daily make a wide variety of decisions, from the easiest (e.g., choosing what to have for breakfast) to the most difficult (e.g., choosing whether to leave a job or what career to pursue). When people make decisions, their perceptions and behaviors tend to show the biases developed by cognitive or social schemas (Augoustinos et al., 2014), such as gender roles. According to Eagly and Wood (2016), gender roles are a prescriptive guide for what women and men are expected to do; women should be concerned about caring for others and men about achievement of their goals. This prescriptive guide seems to lead unconsciously to people behaving and making decisions congruent with their social schema. Indeed, women make decisions directed toward caring for other people, and men make self-oriented decisions, both congruent with social expectations (Fumagalli et al., 2010). These behaviors or decisions are legitimized by society, especially in women, because if they deviate from their traditional gender roles, they could be perceived negatively (e.g., Sutherland et al., 2015). In this respect, recently, Villanueva- Moya and Expósito (2020) showed that women (vs. men) have a greater internalization of gender roles (i.e., femininity), which leads them to make fewer decisions inconsistent with their traditional role due to fear of negative evaluation (FNE). Therefore, for fear of being evaluated negatively, women act in a manner consistent with their traditional role. This pattern can be reflected in family-work decisions, where Dahm et al. (2019) found that women made decisions orientated toward family (e.g., "Let your job be 'secondary' to your spouse/partner's for a period of time" or "Take a job closer to home") and men orientated toward work (e.g., "To travel out of town for work" or "To stay late, go in early, take work home"). Although women make decisions according to what is socially expected, it does not mean they agree with these decisions, because in the long term, they seem to regret prioritizing family, which is negatively associated 1 3 with well-being (Newton et al., 2012). Hence, although men and women have similar decision-making processes, when decisions that can be affected by gender prescriptions to a greater extent are considered (i.e., family and work), they can trigger dilemmas in women themselves, such as longterm regret or fear of rejection, with consequences for their well-being. In this vein, we aimed to understand the meaning of gender and how its internalization seems to be related to sex differences in decision-making and well-being.

Women and Gender Roles
There is empirical evidence both for and against the fact that gender roles have changed today. For example, in the United States, it has been found that gender roles have not changed in recent decades: people have maintained the same beliefs about women and men despite society's progress (Haines et al., 2016). That is, women continue to be regarded as more communal (i.e., warm, friendly, and supportive), and men as more agentic (i.e., assertive, competitive, and independent). In a meta-analysis Eagly et al. (2020) revealed that women have come to be regarded as even more communal over time, but the men advantage in agentic traits showed no change. In Spain, Moya and Moya-Garófano (2021) showed that communal traits were attributed more to women than men over time, and there were no differences in agentic traits. They also found that agentic traits were more associated with women and men in 2018 than in 1985. However, communal traits for women and men were perceived similarly over time. Although there might be a small change in beliefs about gender roles, in general, it can be observed that the traditional role of women (communality) has endured over time. Indeed, women are still considered to be the main caregivers, whereas men are considered to be the main breadwinners (e.g., Cerrato & Cifre, 2018). This reaffirms the social role theory (e.g., Eagly & Wood, 2016), which postulates that women and men's behavior is socially pre-established as a function of gender roles and the characteristics associated with them. Women are socially expected to be communal, consistent with their caregiver role; that is, they should be concerned about caring for others. By contrast, men are expected to be agentic, consistent with their provider role, and therefore they should be concerned about their goal achievements. Both domains (communal and agentic) have been broadly equated with femininity and masculinity self-concepts, respectively (Bem, 1974). Thereby, people integrate femininity or masculinity into their own self-concepts and self-regulate their behaviors according to them. Empirical evidence has shown that sex predicts masculine and feminine gender roles (e.g., Villanueva- Moya & Expósito, 2020;Ward & King, 2018): Women seem to selfreport higher feminine traits, and men seem to score higher masculine traits. Therefore, those who self-report more feminine traits (usually women) are expected to participate in activities related to housework, childcare, or social relations.

Fear of Negative Evaluation, Regret, and Well-Being
Gender roles affect women's behavior to a greater extent than men's behavior because if women's behavior is inconsistent with their feminine traits, they could be socially discriminated, which is a way of limiting women's behavior and therefore maintaining gender inequality (Eagly & Wood, 2016;Rudman et al., 2012). Women are often victims of gender discrimination when they deviate from their traditional role, hence it is not surprising to find that women (self-reporting femininity) have a greater concern about being socially rejected (i.e., FNE) than men, which could cause women to act in a manner consistent with their traditional role (e.g., decision-making; Villanueva- Moya & Expósito, 2020).
FNE refers to the thoughts people experience at the prospect of being judged or viewed negatively by others in social interactions (Leary, 1983). People with higher FNE seem to tend to care more about seeking social acceptance, trying to leave a good impression (Leary, 1983). Although women seem to report higher FNE than men did, some studies have shown that these differences are significant (Biolcati, 2017), while others have not (Gallego et al., 2007;Villanueva-Moya & Expósito, 2020). Considering the aforementioned, experiencing FNE could be a form of negative thinking in women triggered by gender roles, which, in turn could influence their decisionmaking process (Villanueva- Moya & Expósito, 2020). In this sense, FNE has been positively associated with negative thoughts after an event (Makker & Grishman, 2011); that is, individuals with high FNE seem to remember past failures or think about past events more than they want to. This post-event processing could be similar to feeling regret after making a decision. Indeed, Cheek and Goebel (2020) recently found that FNE seems to lead to significant regret about choices. Regret is an emotion associated with making decisions that arises when one realizes or imagines that another choice would have been better than the choice made (Zeelenberg & Pieters, 2007). That is, when person is becoming aware of a better alternative after he or she has made some decision (e.g., as it could be choosing work instead of family for women). This unpleasant emotion has been widely examined in consumption and purchasing decision contexts (e.g., Moyano-Díaz et al., 2014), where sex differences are not usually the focus of study. No sex differences have been reported in regret (e.g., Schwartz et al., 2002), or it has found that women scored higher than men did on regret (van de Calseyde et al., 2018). Despite there being no specific evidence, based on previous literature, we propose that endorsing gender roles could lead to women experiencing regret, through FNE. In many cases, women find themselves in uncertain situations because they would like to progress at work, but at the same time, they feel responsible for the family domain (Campillo & Armijo, 2017), and they could be discriminated if they deviate from it (Eagly & Wood, 2016). Additionally, they would have to choose the domain in which to invest more, because combining family life with employment is problematic for women given that both domains require energy and time, being incompatible with each other in most cases, which can trigger dilemmas (Hochschild & Manchung, 2012). Consequently, they might not be sure if it is better to prioritize a career or family, and they could experience regret about their decisions. Newton et al. (2012) demonstrated that women reported regret about prioritizing family over work or prioritizing work over family. That is, whatever they decided, they seemed to regret it, because both domains are relevant to them. This could be explained by the role of FNE, which could lead women to make decisions based on what is expected of them. On the other hand, considering femininity and regret, Ward and King (2018) found that sex was related to regret through femininity; that is, they found that women (vs. men) self-reported more feminine traits, and consequently, they seemed to regret their decisions. In line with our approach and previous literature, we propose that femininity would lead to higher FNE (Villanueva- Moya & Expósito, 2020), both being mediator variables between sex and regret in decision-making.
Lastly, it should be noted that regret in decision-making has been associated with negative consequences for wellbeing, such as depression, guilt, shame, or lower life satisfaction and happiness (Moyano-Díaz et al., 2014;Schwartz et al., 2002). In this research, we focused on subjective happiness and life satisfaction, both well-being indicators (Diener et al., 2009); according to these authors to obtain a whole picture of an individual's evaluation of their life more than one component must be measured. Given that this research aimed to analyze how the internalization of gender roles could have consequences for women's well-being, both indicators were considered. Additionally, sex differences have been found in these well-being indicators: Men reported higher scores on subjective happiness and life satisfaction than women did (e.g., Batz & Tay, 2018). We propose that the differences in well-being could be explained by the regret experienced by women due to their internalization of gender roles. Considering these variables, we expected that sex would indirectly be related to subjective well-being (subjective happiness and life satisfaction) through femininity, FNE, and regret in decision-making.

Research Overview
Gender roles influence people's behavior, but to what extent might gender roles lead to sex differences in decision-making and well-being? Villanueva- Moya and Exposito (2020) demonstrated that women (vs. men) have a greater internalization of gender roles (i.e., femininity), which seems to limit their decisions through the FNE. We took a step in this direction and analyzed how the FNE not only limits women's behavior but also makes them doubt their decisions.
The current social situation-the progress of women in the public sphere-has given women the opportunity to make decisions related to work (and not only to the family) and progress professionally. Consequently, women (but not men) find themselves in uncertain situations because they would like to progress at work, but at the same time, they feel responsible for the family domain. If the fear of being judged socially is added to this situation of uncertainty-in which women might not be sure if it is better to prioritize career or family-it is expected that they will not feel confident in making their decisions. Hence, regardless of the decision they make, they will regret it, which affects their well-being (Newton et al., 2016). If women made their decisions faithfully, that is, without their decisions being subject to social influences, they would have no reason to regret their decisions, and there would be no sex differences in well-being.
Based on this theoretical conceptualization, in this research, we aimed to understand the meaning of gender and how its internalization seems to be related to sex differences in decision-making and well-being. Specifically, we measured femininity, fear of negative evaluation, regret, and well-being (subjective happiness and life satisfaction). We expected that women (vs. men) would self-report greater feminine traits, lower subjective happiness, and lower life satisfaction (Hypothesis 1). Because empirical evidence is not conclusive regarding sex differences in FNE and regret, we did not expect sex differences in FNE and regret. Additionally, we hypothesized that: femininity would be positively related to FNE (Hypothesis 2); FNE would be positively related to regret (Hypothesis 3); regret would be negatively related to subjective happiness (Hypothesis 4a) and life satisfaction (Hypothesis 4b). Lastly, we examined whether endorsing gender roles could be related to women's decisions (regret), as well as the possible consequences for their well-being. Specifically, we expected that sex (0 = male, 1 = female) would indirectly be related to subjective happiness (Hypothesis 5a) or life satisfaction (Hypothesis 5b) through femininity, FNE, and regret in decision-making.
Finally, we were interested in analyzing sex differences in the most important decisions people make throughout their lives (Hypothesis 6). Women may find both work and family decisions relevant to their life because they would like progress at work but at the same time feel responsible for the family domain (Campillo & Armijo, 2017). We consider that analyzing the content of the most important decisions is relevant because it allows us to determine under what circumstances the proposed model can be contextualized.

Participants
The study sample included 216 Spanish participants (123 women and 93 men; M age = 41.43, SD = 10.89). We excluded thirteen participants because they did not follow instructions. That is, they indicated that they had written about a decision that they were in the process of making instead of writing about a decision they had already made. This allowed us to exclude those participants who did not pay enough attention. The remaining 203 participants (116 women and 87 men) ranged in age from 20 to 65 years old (M = 41.70, SD = 10.93). Most participants (n = 90, 44.3%) had a university degree, followed by high school (n = 71, 35%), secondary education (n = 15, 7.4%), and primary education (n = 11, 5.4%). A majority of participants were employed (n = 135, 64.5%), followed by unemployed (n = 37, 18.2%), housewives (n = 12, 5.9%), and retirees (n = 7, 3.4%). These data do not represent the full percentage because 16 participants did not report both variables. A sensitivity power analysis (G*Power; Faul et al., 2009) for a linear multiple regression test (1 -β = 80%; α = 0.05; N = 203) revealed that the sample size was sufficiently large enough to detect effects of at least a small size of 0.06.

Procedure
Participants were invited to take part in the study through ads placed on social networks (e.g., Facebook) and personal approach. Interested participants were directed to the questionnaire. The inclusion criteria were to be Spanish and to be at least 18 years of age. Before they completed the questionnaire, participants were informed of the general aim ("To analyze possible factors associated with the decision-making process") and the instructions to take part through an online platform. Informed consent was obtained from all participants; they were told about the confidentiality and anonymity of their responses and could decline or agree to answer the questionnaire ("After being informed of the above, I agree to participate in the study"). Once they accepted, they could begin to respond to the measures. Finally, all participants were thanked for their participation and were fully debriefed. This research is part of a broad project approved by the Ethics Committee of the University of Granada. Participation was voluntary, and no monetary incentives were provided.

Femininity
Participants completed the femininity subscale of the Bem Sex Role Inventory (Bem, 1974; Spanish adaptation of López-Sáez & Morales, 1995). Nine items assessed the extent to which people incorporate feminine traits into their self-concepts (e.g., "Sensitive to needs of others"; 1 = never or almost never true, 7 = almost always true). This subscale has shown adequate psychometric properties in Spanish populations (α = 0.73; Villanueva- Moya & Expósito, 2020). In the current study, the subscale demonstrated an adequate internal consistency (Cronbach's α = 0.76).

Fear of Negative Evaluation
Participants completed the Brief Fear of Negative Evaluation scale (Leary, 1983; Spanish adaptation of Gallego et al., 2007). Twelve items assessed the sensation of being evaluated negatively by others. Examples items include "I am usually worried about what kind of impression I make" and "I often worry that I will say or do the wrong thing" (1 = not at all characteristic of me, 5 = extremely characteristic of me). Cronbach's alpha was 0.88 with this sample, similar to the Spanish adaptation (α = 0.90).

Decisions
Participants were asked to "Describe a decision that you have already made, which has a significant impact on your life," which was adapted from Cross et al. (2000). They were encouraged to write in essay style, without a word limit (the first page was used for this). Participants also rated how important the decision was to them ("To what extent was this decision important for you?" 1 = not all important, 7 = very important) and its difficulty ("To what extent was this decision difficult for you?" 1 = not at all difficult, 7 = very difficult). In this way, we wanted to ensure that participants responded to the following measures while thinking of a relevant decision for them.

Regret in Decision-Making
Participants rated five items developed by Schwartz et al. (2002;Spanish adaptation of Moyano-Díaz et al., 2014) to assess the possibility of regretting a decision once made (e.g., "Once I make a decision, I don't look back"; 1 = completely disagree, 7 = completely agree). Cronbach's alpha was 0.67 in the original study, and 0.69 in the Spanish adaptation. In the current study, it was 0.61.

Subjective Happiness
Participants completed the Subjective Happiness Scale (Lyubomirsky & Lepper, 1999), adapted into Spanish by Extremera and Fernández-Berrocal (2014). Four items subjectively assessed the extent to which people self-rated as happy (e.g., "Compared to most of my peers, I consider myself"; 1 = less happy, 7 = more happy). With this sample, the scale showed a Cronbach's alpha of 0.82, similar to the Spanish adaptation (α = 0.81).

Life Satisfaction
Participants completed the Satisfaction with Life Scale (Diener et al., 1985; Spanish adaptation of Atienza et al., 2000). Five items assessed the global judgment that people made about satisfaction with their life (e.g., "In most ways my life is close to my ideal"; 1 = strongly disagree, 7 = strongly agree). The Spanish adaptation obtained a Cronbach's alpha of 0.84. The scale demonstrated an adequate Cronbach's alpha in the current study (α = 0.83).

Sociodemographic Data
We included questions about age, sex, educational level, and employment.

Decision Check
Participants were asked if the decision they had written about was one they had already made or were in the process of making. Participants who indicated that they were in the process of making the decision were removed.

Statistical Analysis Strategy
Prior to carrying out the main analyses, we examined data for normality and multicollinearity (see Table 1). Regarding sex differences, we conducted independent samples t-test analyses using sex as the independent variable and femininity, subjective happiness, life satisfaction, FNE, and regret as dependent variables respectively (see Table 1). Next, we investigated the associations across study variables through correlation analyses (see Table 1). We then carried out two hierarchical regressions analyses to test the initial prediction regarding the effect of sex on well-being indicators through study variables (see Table 2 and 3). We included sex in Step 1, femininity in Step 2, FNE in Step 3, regret in Step 4, and subjective happiness and life satisfaction as the dependent variables, respectively. These statistical analyses were conducted via SPSS (Version 22.0). In order to verify the fitness of the model, a path analysis was performed using maximum likelihood (ML) as the estimator (with 10,000 bias-corrected bootstrap samples; Mplus version 8). The model fit was assessed using the root mean square of approximation (RMSEA), comparative fit index (CFI), Tucker-Lewis index (TLI), and standardized root mean square residual (SRMR), with a RMSEA and SRMR below 0.08, and CFI and TLI above 0.90 indicated good model fil (Hu & Bentler, 1999). Specifically, we examined the indirect effects of sex (0 = male, 1 = female) on subjective happiness or life satisfaction (respectively) through femininity, FNE, and regret. For indirect effects, significance was considered if the 95% confidence interval (CI) from the bootstrap examination did not include zero. Finally, we used the ATLAS.ti 7 software to examine differences in decisions participants had made throughout their lives (see Table 4 and Fig. 2). We first created the hermetic unit and generated the main document containing the decisions literally described by the participants. Second, we codified the decisions and classified them according to the categories of Cross et al. (2000): personal relationships, family, academic, housing, work, or other. To ensure multiple perspectives, the decisions were analyzed and codified by two researchers. Furthermore, we  carried out an independent t-test analysis to analyze sex differences in difficulty and importance of participants' decisions. Table 1 shows descriptive statistics and correlations. As can be observed, there was no problem of multicollinearity because correlations between measures were less than 0.80 (Shrestha, 2020). Regarding normality, the skewness and kurtosis values for all variables ranged from − 0.77 to 0.33, within acceptable limits of ± 2, indicating a normality of distribution (see Table 1; Gravetter & Wallnau, 2014).

Sex Differences in Study Variables
Results showed that women (vs. men) scored higher on all study variables, except well-being indicators (see Table 1). The independent samples t-test analysis showed that, as expected, women (vs. men) self-reported greater feminine traits, lower subjective happiness and life satisfaction (Hypothesis 1). By contrast, the results did not show statistically significant differences in FNE and regret.

Correlations Among Study Variables
As can be observed in Table 1, femininity was positively related to FNE (Hypothesis 2); FNE was positively related to regret (Hypothesis 3). Lastly, regret was negatively related to subjective happiness (Hypothesis 4a) and life satisfaction (Hypothesis 4b).

Indirect Effect of Sex on Well-Being Based on Femininity, FNE, and Regret
Prior to conduct the serial mediation analyses to examine the indirect effects, we carried out two hierarchical regressions analyses prior to test the initial prediction regarding the effect of sex on well-being indicators through study variables. Results showed that the model was significant with subjective happiness, F(1, 198) = 17.42, p < 0.001, with an explained variance of 30% (see Table 2). Likewise, results showed that the model also was significant with life satisfaction, F(1, 198) = 7.08, p = 0.008, with an explained variance of 16% (see Table 3). The path model as displayed in Fig. 1 showed a good fit to the data (RMSEA = 0.00, CFI = 1.00, TLI = 1.00, SRMR = 0.00). The indirect effect of sex (0 = male, 1 = female) on subjective happiness based on femininity, FNE, and regret was significant (b = − 0.02, SE = 0.01, 95% CI [− 0.040, − 0.007]. Thus, Hypothesis 5a was supported. 1 That is, sex (0 = male, 1 = female) was associated with an increase in feminine traits, which was associated with increased FNE. This, in turn, was associated with experienced regret, which seemed to lead to lower subjective happiness. The variables included in the model increased the explained variance of subjective happiness (30%).
Regarding  Thus, Hypothesis 5b was supported (see Fig. 1). 2 Nevertheless, the explained variance of life satisfaction did not increase. 3

Qualitative Analysis of Decisions
Before codifying the decisions, we evaluated the importance and difficulty of the decisions that participants made. These data could reflect the relevance of the decisions to participants. The results showed that, in general, participants considered their decisions important (M = 6.76, SD = 0.63). Thus, the decisions participants described seemed to reflect their most important decisions. No significant sex differences were found in relation to the We then evaluated the occurrences of decision codes. A decision could be codified with one or more codes given that a decision could include multiple categories (e.g., "Leaving paid work to have more time to take care of my family" could include both the work and family categories). In this respect, Poelmans (2005) pointed out that decisionmaking in one domain can be affected by factors in another domain. As can be observed in Table 4 (see also Fig. 2), the most frequent codes for women were work (24%), family (21%), other (20%), personal relationships (18%), academic (11%), and housing (3%). In relation to men, the most frequent codes were work (35%), personal relationships (17%), other (10%), family (9%), housing (9%), and academic (8%).
Next, we carried out an analysis based on the categories. Regarding the code others, decisions included moving to another town or country (4% men and 8% women), pets, and medical treatment, among others. In relation to the code housing, decisions included buying a house (7% men and 2% women) or moving alone (2% men and 1% women). Furthermore, the frequencies of the code personal relationships were similar for both women (18%) and men (17%). Specifically, decisions related to a romantic partnership were the most common (see Table 4). Thus, romantic partnerships seem to be an important part of people's lives.
Furthermore, considering that participants seemed to have reflected on the most important decisions they made in their lives, it seemed that women's decisions were based mainly on work (24%) and family (21%), whereas men's decisions were based on work (35%) to a greater extent. After observing these results, we conducted a deeper analysis with respect to the following domains.
We found that 16% of men's decisions versus 7% of women's decisions were related to changing or leaving a job to improve their working conditions. Woman: "I worked in the family store. Then, the opportunity to start working in an office presented itself. However, the coworkers were men and it was frowned upon for a woman to work. So I did not tell anyone and signed the contract. My father, who I thought would not tolerate it, was glad that I signed without paying attention to what others would think." Man: "I left my job at a company to start my own company. I made this decision because I wanted to grow on the job and my only option was to go to other companies where I had to follow someone else's ideas." Likewise, we found that 1% of men's decisions versus 9% of women's decisions were related to changing or leaving a job to stay with family.
Woman: "I gave up a better paying job to stay with my son. My son was becoming very unstable because of the life we had, since during the day he was in three different places with different rules. The option was to look the other way [reject a better job] and continue in my current job." Man: "I had to leave my job at the bank because they wanted to move me from town and I would have had to leave my family and my house." Lastly, only men (6%) made decisions related to being away from family for work.
Man: "After several years working as a programmer, one day the opportunity came to fill a vacancy at a secondary school. However, it was very far from my house, my daughter, and my family. It would be a risky opportunity, but it would be worth it. That year, I learned a lot and really discovered my true vocation. A year later, I passed the exams, and since then, I have been a teacher. Thanks to that decision, my life took a turn that today I appreciate." It seemed that women's decisions were usually between work and family, prioritizing family. On the other hand, men's decisions seemed to reflect that they did not have to decide between work and family.

Auxiliary Analyses
Regarding the qualitative analysis, discriminant function analyses were used to examine whether categories of decisions could predict participants' regret and well-being as a function of sex. Although the results were not significant (see Supplementary Material), the trends can be seen in the Fig. 3.

Discussion
Gender roles seem to operate as a social schema through which people learn how they must behave; that is, socially, what women and men are expected to do is pre-established (Eagly & Wood, 2016). Gender socialization is such a deeprooted process (Haines et al., 2016) that it is difficult to show its impact on day-to-day issues such as decision-making. Prior research has found that women (vs. men) self-reported more feminine traits, and consequently they experienced more FNE (Villanueva- Moya & Expósito, 2020) and more regret in decision-making (Ward & King, 2018). The current research is the first to examine whether femininity together with FNE could add explained variance of the regret variable, extending previous research. Furthermore, we analyzed the relationship of regret with well-being indicators (subjective happiness and life satisfaction). In this sense, the present research adds to the growing literature on gender roles, showing that they seem to be related to the decision-making process. Specifically, women (vs. men) displayed a greater internalization of gender roles (i.e., femininity), which seemed to increase their tendency to experience higher regret in decision-making because gender roles seemed to increase their FNE. Moreover, this process seems to lead to decreased well-being. Therefore, this research aimed to understand the meaning of gender and how its internalization seems to be related to sex differences in decisionmaking and well-being, perpetuating gender inequality. The current research is framed by social role theory (e.g., Eagly & Wood, 2016), which points out that women are expected to be feminine; that is, women should be concerned about caring for others and men about the achievement of their goals. In line with this theory and previous studies (e.g., Villanueva- Moya & Expósito, 2020), the results of sex differences revealed that women self-reported more feminine traits than men did. In an increasingly egalitarian society, the difference in these traits should decrease; however, this does not seem to be happening. This result seems to indicate that in spite of women's progress in the public sphere, traditional beliefs have been sustained over the years (Haines et al., 2016). Eagly and Wood (2016) argued that women seem to behave according to their traditional role (i.e., feminine) due to the fear of being negatively socially evaluated, given that if they behave nonstereotypically (i.e., in a masculine way), they could be perceived negatively by men and other women (e.g., Rudman et al., 2012). This conceptualization is supported by previous research (Villanueva-Moya & Expósito, 2020) and our study, in which we found that femininity seemed to be positively associated with FNE. Furthermore, in line with previous research, we did not find significant sex differences in FNE; however, women seemed to report higher FNE than men did (e.g., Gallego et al., 2007;Villanueva-Moya & Exposito, 2020). FNE might be triggered by the pressure that women feel if they disregard traditional gender roles (Dinella et al., 2014). Given that, as indicated in the social role theory, gender roles influence women and men's behavior and consequently could limit their behavior (Eagly & Wood, 2016). It should be noted that the evolution of stereotypes has not been the same for women as for men, who are the highest status group and are not judged in the same way as women are, as a subordinate group (Fiske et al., 2002); therefore, women feel more socially pressure.
This fear triggered by feminine roles could influence women's behavior, such as decision-making. In particular, the results showed that FNE seemed to lead to an increase in experiencing regret (Cheek & Goebel, 2020). Furthermore, extending previous research (Ward, & King, 2018), the current research adds FNE together with femininity as a possible mediator variable in the relationship between sex and regret in decision-making. Auxiliary analyses showed that sex indirectly linked to regret in decision-making, through femininity and FNE, such that women (vs. men) self-reported higher feminine traits, which were associated with increased FNE (see Supplementary Material). This, in turn, was associated with more regret. This finding seems to show that gender roles may develop uncertain situations (through FNE) in which women are not sure whether they should make another decision. In this respect, for women (vs. men), the level of FNE triggered by femininity may lead to a decision-making process that is more stressful, leading to greater decision difficulty and thus to experiencing more regret, with the consequent impact of this disruption on their well-being. Indeed, when we analyzed the content of the most important decisions that people made throughout their lives, the findings showed that women perceived their decisions as being more difficult than men did. Gender roles prescribe that, women, should be focused on family (i.e., as caregivers) and men on work (i.e., as providers; Eagly & Wood, 2016). Nevertheless, the increase of women in the workplace has led them to focus on work, too (Campillo & Armijo, 2017), so it is reasonable to think that their decisions are focused in both domains, one for being what is socially expected by gender roles (family) and the other because they want to progress professionally (work). In this sense, as can be observed in the findings of the qualitative analysis, women's decisions are based mainly on the work and family domains, whereas men's decisions are based on the work domain (Hochschild & Manchung, 2012). That is to say, women are encouraged to occupy work positions but without abandoning the family (their gender role) because otherwise they would be socially sanctioned. While men are encouraged to maintain their work role and when they take on family role they are positively viewed for it. This pattern of results seems to reflect that, traditional attitudes are maintained and reinforced; that is, women's decisions are directed toward caring for others as caregivers, and men's decisions are directed toward goal achievement as providers. Therefore, it can be appreciated that the progressive advance toward gender equality in the public sphere (with an increase of women in the work domain) has not spilled over to the private sphere in the same way (with an increase of men in the family domain). Indeed, women continue spending more than twice as much time on family responsibilities than men do (Cerrato & Cifre, 2018). Hence, women may find themselves in uncertain situations because they would like to progress at work, but at the same time, they feel responsible for the family domain (Campillo & Armijo, 2017). This can be seen reflected in the level of difficulty that women perceived in their decisions compared with men given that women have to consider more factors when making a decision. Consequently, women might not be sure if it is better to prioritize work or family, and they might experience more regret. Indeed, in our results, women scored higher than men did on regret (van de Calseyde et al., 2018). Newton et al. (2012) demonstrated that whatever women decided, they reported regret (prioritizing family over work or prioritizing work over family), because both domains are relevant for them (Campillo & Armijo, 2017). Men seemed not to have to decide between work and family, and consequently, their decisions were more focused on achieving their goals (i.e., work), with a lower probability of regret because they did not have to consider other options. Specifically, in the work category, men's decisions were related to changing or leaving a job to improve their working conditions or even to being away from family for work, whereas women's decisions were related to changing or leaving a job to stay with family followed by changing or leaving a job to improve their working conditions. This analysis indicates that women's decisions are between both domains, whereas men's decisions seemed to reflect that their decisions are mainly based on one domain. This dilemma can lead to women being forced to sacrifice one of their options, which can result in feelings of discomfort for the women themselves, such as regret in decision-making. When women are faced with this dilemma, they may choose the family because it is the socially safer option and because women's behaviours are orientated by ideals or expectations (Johnston & Diekman, 2015), but is this really what women want? The simple fact of being a woman and the fear of being judged by others may cause women to decide for the benefit of others (i.e., their traditional role) with the costs that this would entail for their satisfaction. Once the decision is made, with its consequences, women may perceive that the decision they made to sacrifice their career aspirations for the caregiving of others did not benefit them as much as they had hoped, and therefore they regret it.
Previous research has shown that regret in decisionmaking is associated with life satisfaction and subjective happiness (Moyano-Díaz et al., 2014;Schwartz et al., 2002), both well-being indicators (Diener et al., 2009). Researchers have found sex differences in both indicators, with women scoring lower than men (e.g., Batz & Tay, 2018). Could these differences be explained by the regret experienced by women due to gender roles? In the present study, we tested two serial mediation models that depicted sex as indirectly related to subjective happiness or life satisfaction through femininity, FNE, and regret. Specifically, women (vs. men) displayed a greater internalization of gender roles (i.e., femininity), which was associated with increased FNE. This, in turn, was associated with experienced regret in decisionmaking, which seemed to lead to lower levels of well-being (life satisfaction and subjective happiness). Hence, levels of well-being might be explained by the level of uncertainty in women's decisions due to gender roles. In other words, the negative feeling of not knowing if what you have decided is right or wrong because the decision could not be congruent with gender role expectations seems to lead to lower wellbeing in women.
It should be noted that only the indirect effect with subjective happiness as a criteria variable increased the explained variance. Therefore, although the indirect effect was statistically significant with life satisfaction as a criteria variable, caution is warranted in the interpretation of this finding. This could be due to subjective happiness being associated with changes from day to day, whereas life satisfaction is more constant and general (Diener et al., 1997). When an individual makes a decision, the consequences of this decision can change according to daily circumstances and events, which, in turn, can result in regret or not. That is, subjective happiness can change over time, whereas life satisfaction is more consistent. In this sense, the drop in variance in life satisfaction seems to indicate that other possible variables should be considered. Women who are more adherent to gender roles may feel caught between what they want to do and what they should do and may feel regret consistently throughout their lives. Women may feel that this state will not change in the future, given that whatever they choose, they will tend to regret; hence, they do not make an overall negative life appraisal (i.e., life satisfaction). Likewise, it could be said that life satisfaction depends more on what women have achieved throughout their lives and to what extent they have conformed to what is expected (i.e., they behave according to what is socially expected of them in order to feel greater satisfaction despite the regrets they may feel). For example, it is common to hear women say, "I can't complain about my life; I have a family, a job, a good house…," but one might ask whether these women are really happy. Subjective happiness seems to be more circumstantial and variable and, therefore, may vary depending on the choices one makes. For example, the decision to take a six-month leave of absence to care for her baby may affect a woman's happiness (because it is circumstantial), but it may not affect her more consistent perception of her well-being (i.e., life satisfaction). In fact, a woman is socially expected to make this decision because she would be behaving in accordance with the expectations society has established for her.

Limitations and Directions for Future Research
Despite this study's contribution to the gender role and decision-making literature, it has several limitations that are necessary to consider. The study sample cannot be considered representative of Spanish people as it was a convenience sample. Our findings are preliminary, so their generalizability to the general Spanish population must be tested in future research. This research involved cross-sectional data, which limit strong causal conclusions. Future research could complement these findings by using experimental procedures, where the fear of negative evaluation could be manipulated in a sample of women with high femininity. For example, tell participants that they will be given three minutes to prepare a speech to be presented in front of an audience (Durlik et al., 2014). Future researchers might replicate the current findings with other populations or conditions (e.g., research institutes) to analyze these associations in more detail and help define the phenomenon. Based on previous studies (e.g., Villanueva-Moya & Expósito, 2020), we did not assess masculinity. We recommend future researchers consider masculinity as a control variable, given that recent research has found agentic traits to be associated more with women and men nowadays (Moya & Moya-Garófano, 2021). Likewise, it would be interesting to include personality variables. For example, it has been found that women scored higher than men in neuroticism (Murphy et al., 2021), which was positively related to femininity and the FNE (Hazel et al., 2014;Zheng & Zheng, 2011). It might be interesting to control for this variable, because neuroticism assesses the tendency of people to experience negative emotions or related processes (e.g., anxiety) when they perceive threat. Additionally, we recommend that future researchers analyze the degree of regret regarding a decision related to work and/ or family (the most relevant domains) and not regarding decisions in general. For example, researchers could ask participants to describe a decision and indicate their level of regret based on that decision. Furthermore, it might be interesting to ask participants to describe the consequences they perceive of these decisions. Lastly, it would also be interesting to carry out a longitudinal study and to analyze whether gender roles fluctuate or whether women sustain their level of regret.

Conclusion
Gender roles seem to work like a social schema through which women and men learn how they must behave (e.g., decisionmaking). This gender socialization is such a deep-rooted process that people may not awareness of how they have been socialized, which is thus often is difficult to analyze in people's discussions of their own decision-making. On this basis, the current research is one of the first to investigate whether gender roles are related to decision-making factors. Specifically, we investigated whether femininity together with FNE could add explained variance of regret in decision-making. Additionally, we analyzed the possible relationship of regret with well-being indicators (subjective happiness and life satisfaction). The results showed that women (vs. men) had a greater internalization of gender roles (i.e., femininity), which was associated with higher FNE. This, in turn, was associated with experienced regret in decision-making, which seemed to lead to lower well-being. That is, it seems to reflect the relationship between sociocultural variables (gender roles) and individual well-being in women. Specifically, these findings underscore the importance of gender roles to trigger regret in decision-making, which is considered a negative emotion that happens when individuals harm themselves (Berndsen et al., 2004). Overall, this study expands evidence on regret in decision-making through social role theory (Eagly & Wood, 2016), indicating that gender roles can influence women's decisions and well-being, thus maintaining gender inequality. Moreover, the content analysis seems to reflect that although men and women have similar decision-making processes, the content of decisions is determined by the meaning of gender. Therefore, this research tries to understand the meaning of gender and how its internalization seems to increase sex differences in decision-making and well-being.

Ethics Statement
This study was carried out in accordance with the recommendations of Human Research Ethics Committee of the University of Granada.