From 2021... A Tale of Peaks and Valleys: Sinusoid Relationship Patterns Between Mountainousness and Basic Human Values. Stefan Stieger et al. Social Psychological and Personality Science, Aug 16, 2021. https://doi.org/10.1177/19485506211034966
Abstract: Mountains—mythic and majestic—have fueled widespread speculation about their effects on character. Emerging empirical evidence has begun to show that physical topography is indeed associated with personality traits, especially heightened openness. Here, we extend this work to the domain of personal values, linking novel large-scale individual values data (n = 32,666) to objective indicators of altitude and mountainousness derived from satellite radar data. Partial correlations and conditional random forest machine-learning algorithms demonstrate that altitude and mountainousness are related to increased conservation values and decreased hedonism. Effect sizes are generally small (|r| < .031) but comparable to other socio-ecological predictors, such as population density and latitude. The findings align with the dual-pressure model of ecological stress, suggesting that it might be most adaptive in the mountains to have an open personality to effectively deal with threats and endorse conservative values that promote a social order that minimizes threats.
Keywords: personal values, mountainousness, geographical psychology, socioecology, conditional random forests
Discussion
The present research employed advanced analysis techniques to investigate whether mountainousness is meaningfully associated with personal values. Correlation curve analysis indicated that individuals living in hilly and mountainous areas were likely to emphasize conservation values, specifically security and tradition. Individuals living at high altitudes showed a similar pattern but also cared less about hedonism. These results were stable across various robustness checks. Conditional random forest machine-learning algorithms confirmed both mountainousness indices as relevant predictors of personal values when tested against a conservative set of demographic (age, gender, and income) and socio-ecological (population density, latitude) controls.
How should we interpret the associations between mountainousness and personal values? The negative relationship with hedonism appears straightforward. Mountainous areas tend to be secluded and inhospitable, making them ill-suited for the pursuit of worldly pleasures and sensuous gratification. Meanwhile, the robust association between mountainousness and conservation values may initially seem surprising and even counterintuitive. According to voluntary settlement theory (Kitayama et al., 2006, 2010), during the European settlement of the United States, frontier environments like the Rocky Mountains attracted primarily self-reliant, freedom-seeking nonconformists. The accumulation of individuals with such traits laid the foundation for an ethos of independence that continues to characterize the inhabitants of these areas today (Plaut et al., 2002; Varnum & Kitayama, 2011). Indeed, the mountain states still exhibit the strongest individualist tendencies in the United States (Vandello & Cohen, 1999). Moreover, recent research examining the personality structure of mountain dwellers in the United States found that mountainousness was most strongly related to heightened openness to experience (Götz, Stieger, et al., 2020). With openness being negatively related to conservation values (Fischer & Boer, 2014; Parks-Leduc et al., 2015; Roccas et al., 2002), these findings appear to be at odds with the current results.
However, from an analytical standpoint, even the strongest correlations between traits and values—which are typically found between agreeableness and benevolence (rsp = .61, Parks-Leduc et al., 2015; r = .45, Roccas et al., 2002; and r = .54, Vecchione et al., 2019) and openness and self-direction (rsp = .52, Parks-Leduc et al., 2015; r = .48, Roccas et al., 2002; and r = .39, Vecchione et al., 2019)—leave sufficient unexplained variance to manifest in differential relations with third variables, such as mountainousness. More importantly, from a conceptual standpoint, while personality traits and personal values are similar, they are not the same. Values are evaluative, mutually exclusive (i.e., following a diametrical organization, wherein endorsement of certain values implies rejection of others), enduring goals that reflect what a person finds important as a member of society. Meanwhile, traits are descriptive, nonmutually exclusive (i.e., following an orthogonal organization, wherein stronger expression of certain traits does not affect others), enduring dispositions that reflect what a person is like as an individual (Bilsky & Schwartz, 1994; Roccas et al., 2002; Vecchione et al., 2019).
The current findings dovetail well with the dual-pressure model of ecological stress (Conway et al., 2017). According to this model, the same ecological stressor, such as the harshness of mountain terrains, might simultaneously produce opposing pressures that push people in two different directions. In the current context, mastering the tough ecological conditions of mountainous areas might require individuals with independent agency and preparedness to confront unknown challenges and thus favor an open personality (Götz, Stieger et al., 2020). Meanwhile, thriving in ecologically challenging environments, such as mountainous terrains, might require social groups that are committed to safety, self-discipline, stability, and protection of the status quo—hallmarks of conservation philosophy. This conclusion aligns with research showing that experiences of environmental threats and uncertainty (1) prompt individuals to be skeptical of strangers and more territorial about their group domains (Sng et al., 2018), (2) lead to increased endorsement of socially and politically conservative positions (Malka et al., 2014; Oishi et al., 2017), and (3) are conducive to the creation of vertical governmental restriction—laws that impose hierarchies and protect specific groups (Conway et al., 2017, 2020). Thus, having an open personality (i.e., autonomy and the readiness to confront novel challenges when faced with threats) and conservative values (i.e., supporting a social order governed by norms of security, self-discipline and respect for customs to minimize threats) might be most adaptive for thriving in the mountains.3
It should, of course, be noted that the observed effects are small.4 Compared to the average correlation between age and values (M |r| = .098), the average correlation between mountainousness (20 miles) and values was about a 10th (M |r| = .009). However, personal values are determined by many factors (Sagiv et al., 2017), and any single factor is likely to have only a small effect (Götz et al, 2021). This argument is especially true in uncontrolled, real-world settings as in the present study, where—compared to classical lab experiments—effect sizes are typically diminished due to heightened error variance (Maner, 2016; Oishi & Graham, 2010). Moreover, their small magnitude does not render the observed effects unimportant. Rather, even small effects can make a big difference when considered over time and at scale (Funder & Ozer, 2019; Matz et al., 2017). The former seems likely as personal values influence human attitudes and behaviors daily (Sagiv et al., 2017). The latter is especially probable for socio-ecological influences, such as mountains that—while distal and thus less influential than personal factors (e.g., demographics)—simultaneously affect large groups of people who share the same environmental milieu (Conway et al., 2020; Lu et al., 2018; Oishi, 2014). Taken together, the immediate impact of mountainousness on personal values may be small. But when considered over a lifetime and at population scale, small effects translate into highly consequential outcomes such as election results (Caprara et al., 2006), cultural capital, and economic growth (Bardi et al., 2008).
Limitations and Future Research
The current research has several limitations. First, due to the correlational nature of our data, no causal inferences can be drawn. Longitudinal studies at the individual and community levels are needed to illuminate the psychological underpinnings of the associations between mountainousness and personal values (i.e., acculturation effects, selective migration or a combination thereof; Götz et al., in press; Rentfrow et al., 2008; Stieger & Lewetz, 2016). Second, while our data offered one of and perhaps the largest personal values samples in the United States, it is not nationally representative. Although the ethnic composition and geographic coverage were broadly representative of the general population, which is common in large-scale online samples (Gosling et al., 2004; Götz, Bleidorn, et al., 2020; Jokela et al., 2015; Kosinski et al., 2015), the participants in our study were younger, predominantly female, and less affluent than the national average (U.S. Census Bureau, 2020). Third, our assessment of personal values was limited to a 20-item short scale. While the TwIVI displayed respectable psychometric properties in the current study and previous research (Sandy et al., 2017; Vignoles et al., 2018), its brevity comes at the cost of reduced measurement precision and content breadth (Credé et al., 2012). Thus, future research should extend the current work by using longer scales, which might include the extended 19-value version (Schwartz et al., 2012) that could offer even more nuanced insights. Such work may also systematically assess nonlinear trends in mountainousness–value associations (Lee et al., 2021).5 Furthermore, future research might try to dynamically adjust the 20-mile radius as a proxy for the mean commuting distance to the actual commuting distance in each ZIP-code area. Such an adjustment might reduce error variance and isolate the effect of interest more effectively. Lastly, future research should investigate the associations between personal values and other challenging ecologies, including coastlines, swamplands, and deserts (Götz, Stieger, et al., 2020; Oishi et al., 2015).