Do Authentic People Care about the Environment? A View from Two Paradigms

Background Personal authenticity, the ability to be true to oneself, is traditionally studied from the perspective of its protective role for the individual and is only beginning to be studied in relation to the surrounding world. In this study, we suggest that authentic people may be more aware and concerned about their environment then less authentic people. The theoretical foundations for our work were: the person-centered approach; subject psychology; and modern research on pro-environmental behavior. Objective We presented our understanding of personal authenticity within Russian subject psychology, developed the standardized instruments necessary for carrying out our main aim, and explored the links between authenticity and pro-environmental behavior in both person-centered and subject psychology. Design Four hundred thirty (430) Russian students (Mage = 19.19; SDage = 1.22; 79.5% women) participated in the study. Authenticity was measured both bythe revised Russian version of the Authenticity Scale, and a new tool, the Moscow Authenticity Scale, which was developed on the basis of subject psychology. To measure pro-environmental behavior, we created a new instrument called the Ecological Lifestyle Scale, which included Social Activities and Ecological Self-restraint subscales. Results Using the two new scales, the Moscow Authenticity Scale and the Ecological Lifestyle Scale, along with a modification of the Authenticity Scale, we found that authenticity, considered within the framework of subject psychology, provided a more nuanced picture of the relationship between personal authenticity and pro-environmental behavior than the person-centered model did. Women were more likely to exercise pro-environmental behavior than men; however, the connections between personal authenticity and pro-environmental behavior were stronger in the male group. Conclusion Authenticity is associated with pro-environmental behavior but does not predict it accurately enough. Future research on moderating or mediating variables is suggested.


Introduction
. ese proposed changes are often viewed as an educational objective in schools and universities (Zsóka, Szerényi, Széchy, & Kocsis, 2013), based on evidence that younger people are more adaptable in their attitudes (Lie änder & Bogner, 2014) and are more focused on and motivated toward building a constructive relationship with nature.
One idea of an ecologically-minded person is represented in Carl Rogers' vision of future human moral values. In Rogers' view, human beings have an intrinsic potential to care for nature and be close to it (Rogers, 1995). Since his work, environmental psychology has found several nature-related phenomena that con rm his ideas of a deep link between humans and the natural world (Clayton, 2003;Mayer & Frantz, 2004). However, this link seems to be weakened by the overwhelming stress and monotony of modern life, and a lack of time spent in natural surroundings, which increases the risk of developing psychological disorders like the nature de cit disorder (Kuo, 2013). Connection to nature also appears to be gender-related. Women have consistently shown higher relatedness to nature (Lawton, Brymer, Clough, & Denovan, 2017;Irkhin, 2020) and more participation in pro-ecological activities than men (Richardson, Cormack, McRobert, & Underhill, 2016;Dietz et al., 1998;Kennedy & Kmec, 2018).
Taking into account the rather unique Russian culture and mentality, o en placed somewhere in between collectivism and individualism (Mamontov, Kozhevnikova, & Radyukova, 2014), in Russia a new model of authenticity called subject psychology was developed by the famous Soviet psychologist Sergei Rubinstein (1889Rubinstein ( -1960. Rubinstein was one of the "pillars" of Soviet psychology. He was educated in Germany and demonstrated an unusual breadth of interests and dialectical thinking. According to his theory (Rubinstein, 2012), which was in uenced by the works of Franz Brentano, personality always exists in the context of the environment, circumstances, and the world as a whole. We cannot extract an individual from this context; thus, the true self necessarily includes the presence of the social and natural circumstances of their life.
In contrast to Rogers' person-centered approach, which considers social in uences as a source of violation of the individual's authenticity, the subject approach sees the true self involving mutually productive interactions of the individual with other people, culture, and nature (Znakov, 1998;Znakov, & Sverchkova, 2003). ese ideas were taken into account when we developed the Moscow Authenticity Scale, presented below.
Authenticity, de ned as the ability to be oneself or to follow one's true self, is considered in our study as the coherence of a person's life experiences (actions, cogni-tions, and emotions), on the one hand, and his/her personality (temperament, values, beliefs) and the circumstances of his/her life (time, place, and life-calling), on the other (Nartova-Bochaver, Reznichenko, & Maltby, 2021). In contrast to the personcentered model of authenticity, which has been criticized as an individualistic one (Strohminger et al., 2017), authenticity in subject psychology acts as an integral characteristic of a person, represented by both a subjective attitude toward the self and the world around them (Slobodchikov & Isaev, 1995). us, authenticity appears as a holistic phenomenon, implying the inseparability of the person from the world, and their interdependence. In the prism of subject psychology, the acquisition of one's true self is carried out through ethical attitudes and placing oneself in the context of social relations (Rubinstein, 2012). erefore, involvement in social relationships and the establishment of honest, harmonious relationships with others, are inseparable from authentic living and self-realization.
Ottiger and Joseph (2020) revealed a signi cant positive connection between authenticity and ethically minded consumer behavior, providing the rst empirical support for Rogers' idea of an ecological mindset. However, when taking into account that the person-centered conceptualization of authenticity does not work very well in Russia , it is necessary to consider an alternative understanding of authenticity, namely, one developed within the framework of Russian subject psychology.
In the current research, we examined the link between authenticity (understood from the two points of view) and pro-environmental behavior. We considered proenvironmental behavior as those actions that increase environmental bene ts or reduce environmental harm (Steg & Vlek, 2009).
To investigate the possible link between personal authenticity and pro-environmental behavior, we hypothesized that: 1) Personal authenticity would positively correlate with pro-environmental behavior; and 2) This connection would be moderated by conceptualization (in the framework of subject psychology, it would be stronger than with the person-centered approach).
First, we present the preliminary psychometric work which was required to perform our main study.

Participants
e students took this survey as part of their homework, via the online service 1ka.si. A total of 430 people, in the age range of 17-26 years old (M age = 19.19; Me age = 19.0; SD age = 1.22; 79.5% women) participated. All participants were bachelor or master's program students from Moscow universities. Participation was voluntary; all participants gave their informed consent to anonymously publishing their data. Along with the main questionnaires, respondents provided demographic information (age, sex, ethnicity, and religion).

Measurement Instruments
Authenticity Scale. e modi ed Authenticity Scale measures an individual's selfreported authenticity trait. Based on the person-centered approach, authenticity is conceptualized as a tripartite construct comprising Authentic Living, not Accepting External In uence, and a lack of Self-Alienation (Wood et al., 2008) (Appendix 1). It has 11 items, with a seven-point Likert scale, ranging from 1 (does not describe me at all) to 7 (describes me very well).
Moscow Authenticity Scale (MAS). e MAS is a new one-factor tool developed within the framework of subject psychology. It consists of ve items (Appendix 2), with a ve-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree).
Ecological Lifestyle Scale (ELS). e ELS is a new tool developed to measure proenvironmental behavior as a person's stable behavioral pattern, re ecting their ecocentric worldview. e ELS includes two subscales: Social Activities (three items), which describes the person's purposeful social activities aimed at protecting nature, and Ecological Self-Restraint (four items), which describes routine pro-ecological actions aimed at an environmentally-friendly lifestyle. A ve-point Likert scale, ranging from 1 (never) to 5 (very o en), was used (Appendix 3).

Analytical Strategy
First of all, item pools were developed for two new questionnaires, the Moscow Authenticity Scale and the Ecological Lifestyle Scale. e items were developed based on both deductive (literature review and assessment of existing scales) and inductive (exploratory research methodologies, including focus group discussions and interviews) methods. Seven experts in personality psychology and environmental psychology participated in formulating the questionnaire items.
A er the experts' deliberation, item reduction, dimensionality testing, and verication of psychometric properties (model quality and reliability) of the new questionnaires were carried out. Decisions on the reduction of items and the choice of the optimal factorial structure of the questionnaires were based on the results of Horn's parallel analysis (both principal axis analysis and principal components), exploratory factor analyses (EFA), and con rmatory factor analyses (CFA). Both EFA and CFA were performed using the robust ML (MLR) rescaling-based estimator, due to its ability to handle ordinal variables. To assess the model t of each of the developed questionnaires, we examined both absolute and incremental t indices, including the Root Mean Square Error of Approximation (RMSEA) and its 90% con dence interval, as well as the p-value of Close Fit (PCLOSE), the Comparative Fit Index (CFI), the Tucker Lewis Index (TLI), and the Standardized Root Mean Square Residual (SRMR). e internal reliability of all developed scales was estimated with both Cronbach's Alpha and McDonald's Omega.
Finally, the relationships between authenticity (the independent variable) and pro-environmental behavior (the dependent variable) in male/female groups were examined via ANOVA, the Pearson correlation, and multiple regression analysis. To address the imbalanced classi cation problem (20.5% male), we used weighted sampling while processing regression analyses (weights of 3.5 for the minority class) and also used bootstrap, based on 1000 samples, to check the signi cance level of predictors in the regression model.
Statistical analysis was performed using R So ware and Programming, environment 4.0.2 (R Core Team., 2020), SPSS v. 23, and MS Excel 2016 so ware.

Modi cation of the Authenticity Scale
A recent study  had presented a Russian version of the Authenticity Scale (Wood et al., 2008). Although it was a working instrument, we decided to modify it, because the authors reported a bias towards lower scores, a potential ceiling e ect, and the cognitive complexity of one item. Moreover, the subscale measuring Authentic Living in this version, turned out to be inverted relative to the original, and we hoped that by selecting more accurate wordings, we would be able to make it direct; unfortunately, this was not possible.
In the current study, we developed additional ("spare") versions for ve items; the resulting scale included 16 items. All previous studies devoted to the analysis of the factor structure of the Authenticity Scale (Wood et al., 2008;Di Fabio, 2014), indicated that the hierarchical model (three uncorrelated rst-order factors and a higher-order Authenticity factor) describes empirical data better than a rst-order correlated model or a bifactor one. Taking into account the high-level reproducibility of the results and reliability of the hierarchical model of the Authenticity Scale, and also small substantive changes in the current version of the scale, we focused on the analysis of the hierarchical model and did not compare its t indices with other alternative models. Based on the CFA modi cation indices, factor loadings, and data on item distributions, we chose the optimal three-factor model, which consists of 11 items ( Appendix 1). As expected, the t indices and internal reliability of the modi ed version of the model were better than in the model created during primary validation (Hu & Bentler, 1999) (Table 1). All items were normally distributed (ranging from -1 to 1); the factor loadings o f the current model were also higher than the loadings obtained in the primary validated model: 0.57-0.91 vs. 0.53-0.79 on the rst-order factors, and 0.71-0.92 vs. 0.72 to 0.89 on the second-order authenticity factor. e modi ed model explained 68.9% of the variance, while the primary validated model explained only 56.7%. As in the original version, the current one includes three subscales, such as Authentic Living, not Accepting External In uence, and a lack of Self-Alienation.

Development of the Moscow Authenticity Scale (MAS)
Item pool development We aimed to design the MAS as a time-e ective express tool which should measure personal authenticity as a holistic phenomenon in line with subject psychology. When formulating the MAS items, we emphasized such features of an authentic person as the acceptance of when, in the stream of time, a person has been "thrown-in" to life, the coherence of one's personality, and one's general life course, in accordance with our de nition of authenticity. us, the MAS items re ected a synergistic, holistic model of a person-in-the-world. In contrast with the items of the Authenticity Scale, interaction with the social world was interpreted in terms of environmental pressure, and most wordings described manifestations of self-alienation, rather than authenticity itself.
Seven experts worked together to develop and re ne the items and consulted other scholars. e pool included 12 items; then, we set a limit of 5-7 items as the target number of statements.
Factor analysis and reliability testing A principal component exploratory factor analysis, with varimax rotation and extraction based on eigenvalues greater than 1, was conducted, from which a two-factor solution (accounting for 58.25% of the variance) converged in four iterations. Five items, with loadings lower than 0.50 and/or cross-correlations above 0.30, were dropped and an additional iteration was conducted, which resulted in a one-factor structure (explained variance = 55.30%), with loadings for the remaining items higher than 0.50. e results of Horn's parallel analysis showed that a one-factor solution was optimal.
Further, the resulting one-factor model, with the remaining seven items, was tested using CFA. e scaled t indices (χ 2 (14) = 63.773; p < 0.001; CFI = 0.945; TLI = 0.917; RMSEA = 0.090 (95% CI [0.069; 0.114]); and SRMR = 0.410) were unacceptable (Hu & Bentler, 1999). To make an additional improvement of the model, we checked the modi cation indices (MI) with values higher than 10. e MIs suggested that the model could be improved by drawing error covariances between the items "I live in accordance with my beliefs" and "Although I'm wrong, I'm living my own life") (MI = 23.06), and the item "I know for a fact that I am not living my life in vain, " along with three other items (MI = 11.15-15.67).
Since there was no strong theoretical rationale to add the error covariances between the items, we decided to remove the two items with the lowest factor loadings ("I live in accordance with my beliefs" (0.50) and "I know for a fact that I am not living my life in vain" (0.51). ese changes improved the model: all t indices showed a perfect t of the model (χ 2 (5) = 3.767; p < 0.583; CF = 1.000; TLI = 1.000; RMSEA = 0.000 (95% CI [0.000; 0.051]); PCLOSE = 0.946; and SRMR = 0.017) (Hu & Bentler, 1999). e range of the factor loadings of the ve items included in the model was 0.51-0.78; the amount of variance explained by the model was 51.8%.

Development of the Ecological Lifestyle Scale (ELS)
Item pool development To accomplish the main aim of the study, we needed a new instrument that would measure a person's pro-environmental behavior and ecocentric belief system, since most of the existing tools were developed in WEIRD (Western, educated, industrialized, rich, and democratic) cultures, which di er from the lifestyle of Russians.
During the development of the item pool, some statements were borrowed from works by Markle (2013) and Clayton et al. (2021) and subsequently modi ed (questions were reformulated into statements according to forward and back-translation guidelines ("Process of translation and adaptation of instruments, " n.d.). As a result, the initial pool included 12 items, re ecting both environmental routine actions (e.g., water and power conservation, sorting of waste, reduced car use, refusal of meat products, etc.) and civic actions (e.g., prosocial behavior; volunteering for nature protection activities, and donating to environmental charities).

Factor analysis and reliability testing
According to Kaiser's eigenvalue > 1 criterion, the number of factors to be retained was three (eigenvalues of 4.36, 1.47, 1.11), which accounted in total for 57.74% of the variance, but examination of the scree plot suggested only two factors. Based on Horn's parallel analysis, a two-factor solution was optimal. During EFA, one item ("If possible, I avoid consuming animal-based food [meat products]") was excluded due to its low factor loading (> 0.3).
When conducting CFA, we built a hierarchical model with two uncorrelated rstorder factors and a higher-order factor -Total Ecological Lifestyle. Our choice of the higher-order model, rather than the correlated factors model, is explained by the fact that a person's pro-environmental behavior is considered as an integral construct consisting of several subordinate "traits" (Markle, 2013;Clayton et al., 2021). Accordingly, the structural model should assume the possibility of calculating scores on both the subscales and the overall score, unlike the correlated factors model, which can't incorporate any general factor. When investigating a new tool, Brown (2015) suggests using the higher-order structure rather than the bifactor model to explore theoretical understandings of the relationship between a series of subscales which are distinct from one another but united by a common factor. e model t values were not satisfactory; however, the two latent factors loaded highly (< 0.6) on a higher-order factor, suggesting that a hierarchical structure was appropriate to explain the relationships between the latent variables. To improve the quality of the model, four items with factor loadings below 0.50 and/ or high error covariances between the items were removed (Appendix 3). e resulting hierarchical model, with two latent factors, showed satisfactory scaled t indices: (χ 2 (12) = 11.013; p < 0.528; CFI = 1.000; TLI = 1.000; RMSEA = 0.000 (95% CI [0.000; 0.044]); PCLOSE = 0.977; and SRMR = 0.024) (Hu & Bentler, 1999). e rst factor, labeled Social Activities, contained three items (participation in voting, rallies, and house meetings). is subscale had satisfactory reliability values: ω = 0.828 and α = 0.812 (Hair et al., 2010). e second factor, named Ecological Self-Restraint, included four items regarding actions aimed at consumption reduction and a nature-friendly lifestyle. e reliability of this scale was acceptable (ω = 0.675; α = 0.643) for a research instrument with few items (Taber, 2017).
e Total Ecological Lifestyle factor explained 0.807 of the variance at the rstorder factor level (hierarchical Omega); Cronbach's Alpha was also satisfactory (α = 0.746). e factor loadings on the rst-order factors were reasonable (between 0.43 and 0.86). Taken together, Social Activities and Ecological Self-Restraint explained 59.7% of the variance and loaded highly on a higher-order Total Ecological Lifestyle factor (0.63 and 0.79 respectively).

Main Study: the Connection between Personal Authenticity and Pro-environmental Lifestyle
First of all, we were interested in whether there were any gender di erences between the means of the investigated variables. Since the assumptions of normal univariate distribution (the values for asymmetry and kurtosis were between -1 and +1) and the equality of variances (Levene Statistic > 0.05) were met, an ANOVA was carried out, which demonstrated that men and women signi cantly di ered in the mean values of Accepting External In uence (F(1; 429) = 7.36; p = 0.007); Social Activities (F(1; 429) = 10.66; p = 0.001); Ecological Self-Restraint (F(1; 429) = 7.15; p = 0.008); and Total Ecological Lifestyle (F(1; 429) = 12.61; p = 0.000) ( Table 2).
Next, we analyzed the correlations between authenticity and ecological lifestyle variables (Table 3). Weak, but signi cant correlations were found between the Moscow Authenticity Scale (r = 0.166; p = 0.000), Authentic Living (r = 0.100, p = 0.005), and Self-Alienation (r = -0.121; p = 0.001 respectively), on the one hand, and the Social Activities subscale on the other. Ecological Self-Restraint had no interrelations with the authenticity variables. Total scores on the Ecological Lifestyle Scale positively correlated with the Moscow Authenticity Scale (r = 0.136, p = 0.001) and Authentic Living scores (r = 0.083; p = 0.022) and had a negative relationship with the Self-Alienation subscale (r = -0.094; p = 0.009). e correlations between ecological lifestyle (ELS) variables and the person-centered tool of measuring authenticity (MAS) were stronger than the correlations obtained between ecological lifestyle (ELS) and the person-centered framework of Authenticity consideration (the Authenticity Scale), con rming Hypotheses 1 and 2. Note. Signi cant di erences in mean values on subscales for men and women are in bold (p < 0.05). AL = Authentic Living; AEI = Accepting External In uence; SA = Self-Alienation; MAS = the Moscow Authenticity Scale; SoAct = Social Activities; Sf-R = Ecological Self-Restraint; ELS = total scores on Ecological Lifestyle Scale Table 3 Correlations between MAS, Authenticity Scale, and ELS Multiple regression analysis was carried out, using the enter method, in order to determine whether authenticity could predict pro-environmental behavior. ree subscales of the Authenticity Scale, and the scores on the Moscow Authenticity Scale, as well as gender (dummy-coded: 1 = male, 2 = female), were taken as independent variables, and the Social Activities and Ecological Self-Restraint subdomains of ecological lifestyle were taken as the dependent variables. e regression model for Social Activities and Ecological Self-Restraint included two signi cant predictors: scores on the Moscow Authenticity Scale and gender ( Table  4). Although technically both models were signi cant, con rming connections be-tween the investigated variables, the determination coe cients of both models were low (R 2 = 0.073 and 0.027). e scores on the Moscow Authenticity Scale and gender explained the variability level of ecological lifestyle subscales poorly, amounting to less than 8%. us, we cannot state that authenticity signi cantly contributed to an ecological lifestyle. At the next stage, we analyzed the links between the authenticity variables and ecological lifestyle subscales in the female and male groups separately. According to the results of the regression analysis, only the MAS scores made a positive and signicant contribution (β = 0.163; p = 0.002) to Social Activities in the female group. Ecological Self-Restraint was not associated signi cantly with any of the authenticity variables ( Table 5). e signi cant predictor of Social Activities in the male group were No variables were entered into the equation

Discussion
Before we could prove the main hypotheses of the study, we had to develop or modify our techniques: the renewed versions of the Authenticity Scale (Wood et al., 2008), the MAS, and the ELS. e Russian version of the Authenticity Scale (Nartova-Bochaver et al., 2021) went through a set of adjustments. Several items were reformulated for better clarity, and ve additional versions of the translation were needed to replace the previous ones, which led to better internal reliability and t indices. Unfortunately, we were not able to keep the Authentic Living subscale not inverted as in the original scale.
We also developed a new short tool for assessing an authentic personality, named the Moscow Authenticity Scale (MAS), in light of the fact that subject psychology developed as a scienti c school mainly in Moscow. e content of the subscale items re ected the main idea of subject psychology, which studies not the person and the world but the person-in-the-world. is also corresponds to the collectivistic side of Russian culture. e Moscow Authenticity Scale consisted of ve items and one factor; the scale showed acceptable internal consistency and a perfect t to the empirical data.
We also created the Ecological Lifestyle Scale (ELS), which measured pro-environmental behavior in the context of an impact-oriented approach. is scale was developed with regard to the pro-environmental opportunities, culture, and lifestyle of Russians. It consisted of seven items and had two factors, labeled Social Activities and Ecological Self-Restraint. e scale showed good psychometric qualities: t scores and internal consistency. e Ecological Self-Restraint subscale's reliability was slightly below the accepted range, but was still acceptable for an instrument with few items, and created purely for research (nonclinical) purposes (Taber, 2016).
Our main goal was to study the relationship between personal authenticity, understood in terms of the two research paradigms, and pro-environmental behavior.
We found a weak but signi cant correlation between both measures of authenticity and one of the two subscales (Social Activities) of the ELS. e MAS scores showed a strong positive correlation with Social Activities, whereas the corresponding subscale Authentic Living of Wood's Authenticity Scale, did not form any connections at all. is demonstrated that an authentic lifestyle in the framework of person-centered psychology is absolutely orthogonal to the ecological lifestyle. In other words, among authentically living people, there could be both pro-ecological and anti-ecological ones. However, reverted subscales of the Authenticity Scale, Accepting External In uence, and Self-Alienation, formed clear negative connections with Social Activities: people who experience environmental pressure and those who are alienated from themselves, are not inclined to support an environment that is not friendly to them.
Interestingly, Ecological Self-Restraint did not correlate with any measures of authenticity; the question arises: why is this so?
In our opinion, the items of the Social Activities subscale clearly re ect the interdependence of a person with other people (s/he encourages other people to be active) and with nature (this activity is aimed at protecting the environment), which fully corresponds to understanding personal authenticity within the framework of subject psychology. By contrast, the items on the Ecological Self-Restraint subscale are not so unambiguous and allow, along with environmental motivation, pragmatism (for example, saving money for paying for electricity), which may not relate to the "true self " in any way.
In addition, the items of the rst subscale are obviously formulated in the "promotion" modality, whereas the items of the second one are formulated in the "prevention" modality. An authentic person, in accordance with the classical understanding, is creative and capable of productive activity. It can be assumed that, if pro-ecological items were formulated as motivational (for example, to clean one's yard, plant a tree), then connections would be obtained. At the moment, this is speculation, but it can serve as a starting point for further research.
Based on these results, we partially con rmed our Hypotheses 1 and 2, with some quali cations: higher authenticity was connected with social activity but was not connected with conservation behavior. e results partially overlap with those obtained by Ottiger and Joseph (2020).
In the main study, we also witnessed ambiguous results based on gender. First of all, Accepting External In uences, and the totals of the Ecological Lifestyle, Social Activities, and Ecological Self-Restraint scores were lower in the male group. is means, in line with previous results (Irkhin, 2020;Lawton, Brymer, Clough, & Denovan, 2017), that the men were less closely connected with the outer world, compared with women. In other words, men seemed to be more individualistic (Borkenau, McCrae & Terracciano, 2013). e next step of analyses we performed was the development of a multi-regression model. e only predictors of Social Activities (a subscale of the ELS) were gender and authenticity (MAS); Ecological Self-Restraint was predicted only by gender. Although the pro-ecological variables' level was lower in males, men's authenticity showed more connections with pro-ecological variables than women's did. Two regression models of authenticity measures, aimed at predicting pro-ecological behavior, were built separately for the two gender groups, showing mixed results. In women, Social Activities were predicted by authenticity, measured by the MAS, while there were no predictors for Ecological Self-Restraint. In men, MAS positively predicted Social Activities and Ecological Self-Restraint (both ELS subscales), while Authentic Living (a subscale of Wood's model) predicted Ecological Self-Restraint negatively.
us, these outcomes also con rmed Hypothesis 2. ese results re ect an essential di erence between the two models of authenticity. According to Rogers' person-centered model, authenticity, even in the case of a morally mature person, cultivates independence from the outer world, including connections with nature and the environment. So, authenticity might contrapose self-transcendence, a value that is consistently associated with environmentalism (Lee, 2018). Indeed, if one's motivation is to unite with nature, one is required to let go of one's uniqueness and thus, in certain sense, one's authenticity. So, when pro-environmental action requires a sacri ce of personal resources, an individualistic person encounters a contradiction between their own needs and the needs of a sustainable society. Resolving this contradiction egoistically (valuing their own needs over the needs of others) would be authentic for an individualistic person.
On the other hand, the subject psychology approach, promoted in the MAS, implies congruence between a person and society, therefore presuming that the values of a sustainable lifestyle are a part of one's true self in the form of one's larger purpose in life. is ideally allows one to avoid the contradiction between personal needs and the needs of a sustainable society, which, in our case, means that ecological behavior is not experienced as a sacri ce, but a means to ful ll a greater ecological goal.
Overall, the MAS showed higher correlations with an ecological lifestyle and was better at predicting pro-ecological behavior than the subscales of the Authenticity Scale of Wood et al. (2008). But in general, all the regression models had low coefcients of determination, which means that both authenticity measures were unable to accurately explain the variability of pro-ecological behavior.

Limitations and Future Research
e current study was conducted with Moscow university students; our sample was also gender-biased (79.5% women), which further limits our ability to extrapolate the results to a larger population. Apart from extending the age range of the sample, it would be interesting to obtain data from people living in small towns and rural areas in order to examine the results with people with a more diversi ed experience of nature. Moreover, as our results are not easily interpreted, it would be worthwhile to consider other variables, and to build more complicated models, including moderating and mediating e ects. We should also investigate the correlations between speci c items in the ELS, to see which particular pro-environmental activities would correlate with authenticity. Another challenge that we set ourselves is to conduct and describe a more complete validation study of the MAS and ELS, including crossvalidation and measurement invariance across age and gender. Finally, we can use (or develop) more instruments to measure environmentally friendly attitudes and behaviors.

Conclusion
e results of the current study provide some insight into the growing research on personal authenticity and its bene ts in relation to the environment. While other research has suggested a connection between authenticity and pro-environmental values (Ottiger & Joseph, 2020), the current research brought mixed results. Generally, authenticity is associated with pro-ecological behavior. However, it does not accurately predict it, which motivates us for future study of the connection, in search of possible mediators.
In comparison to Wood's Authenticity Scale (2008), the MAS showed stronger correlations with an ecological lifestyle, especially with the Social Activities score, and therefore is proposed as a more accurate instrument in light of the speci c Russian mentality and cultural nuances.
Results of the current study indicate that women are more likely to exercise proecological behavior than men (while showing no signi cant di erence in authenticity between genders), but the connections between authenticity and pro-ecological style were more nuanced in males.
As a side but very signi cant outcome, the current paper introduced three freshly developed scales necessary to conduct the main study: the revised Russian version of Wood's (2008) Authenticity Scale; the Moscow Authenticity Scale; and the Ecological Lifestyle Scale. ese scales show good psychometric properties and are valid for measuring authenticity and pro-environmental behavior in a Russian youth sample and can be recommended for use in various areas of non-clinical research and practice.

Ethics Statement
is research was approved by the Commission for the Ethical Evaluation of Empirical Research Projects of the Department of Psychology of the NRU HSE.

Author Contributions
S. N.-B. developed the theory, supervised the ndings & data collection, and wrote a dra ; S.R. and B.I. developed the method, performed an empirical study and data analysis; S.R. performed computations and wrote a dra . B.I. wrote a dra , edited the manuscript, and revised the paper a er reviewers' suggestions. All authors discussed the results and contributed to the nal manuscript.