The biophilic work design model applies Kaplan and Kaplan’s (1989) seminal attention restoration theory to workplace settings to explain how contact with nature at work can have energizing effects on employees. By integrating theories of human energy (e.g., Kruglanski et al., 2012; Quinn et al., 2012), with research on nature’s impact on emotional, cognitive, physical, and prosocial outcomes (e.g., Keniger et al., 2013), Klotz and Bolino (2021) identify how greater contact with nature at work can restore employees’ potential energies; resources that can be stored and drawn upon in the future (Quinn et al., 2012). Specifically, emotional potential energy supports positive feelings (e.g., enthusiasm) and helps employees regulate their emotions, cognitive potential energy allows employees to regulate their thoughts and maintain directed attention, and physical potential energy enables feelings of health and strength. Each of these forms of potential energy can be replenished by contact with nature (e.g., Kaplan and Berman, 2010; Hartig et al., 2014; McMahan and Estes, 2015; de Keijzer et al., 2016; Twohig-Bennett and Jones, 2018).

However, Klotz and Bolino (2021) focus on energies and do not identify specific work-related outcomes in their theory. Therefore, we extend the biophilic work design model by examining job engagement and creativity at work. Job engagement is defined as a positive motivational state characterized by the exertion of emotional, cognitive, and physical energy in one’s work role (Rich et al., 2010). Emotional engagement is characterized by high pleasantness and activation, including feelings of positivity, enthusiasm, and interest in one’s job. Cognitive engagement reflects focus and concentration in one’s job. Physical engagement is defined as working with intensity and exerting energy in one’s job. Next, creativity is defined as the generation of new and useful ideas, approaches, or solutions (e.g., Hennessey and Amabile, 2010; Anderson et al., 2014; Mumford and Todd, 2019). Like engagement, creativity at work is considered a motivational construct (e.g., Amabile, 1997; Zhou and Shalley, 2011) and relies on available emotional, cognitive, and physical energies to perform job tasks that may require or benefit from creativity. For instance, positive affect and emotions are conducive to creativity at work (e.g., Amabile et al., 2005; George and Zhou, 2007), being creative is an inherently cognitive process that requires higher-order cognition (e.g., Dietrich, 2004), and feeling energized at work is associated with workplace creativity (e.g., Atwater and Carmeli, 2009). In this way, energies gained by exposure to nature should enhance employees’ job engagement and creativity at work.

Additionally, as another extension of the biophilic work design model, we consider the built and natural environment in both the work and nonwork (i.e., home) domain. Specifically, the type of outdoor environment where individuals work and live are explored, and their experiences of time spent outside during work and nonwork time are examined. Ultimately, the replenishment of energies may be different across these unique settings, resulting in different practical implications about where, when, and how time should be spent outside to foster engagement and creativity at work.

We explored associations among the built and natural environment where individuals work and live, experiences of being outside, and work-related outcomes. The limited studies on nature in relation to job engagement and creativity have either exclusively focused on the indoor work environment (e.g., office plants, window views, videoconferencing backgrounds) or examined time spent outside more generally, rather than considering the surrounding built and natural environments (Bringslimark et al., 2007; Nieuwenhuis et al., 2014; Plambech and Van Den Bosch, 2015; Korpela et al., 2017a,b; Hyvönen et al., 2018; Thompson and Bruk-Lee, 2019; Hähn et al., 2020; Fleury et al., 2021; Petersson Troije et al., 2021; Palanica and Fossat, 2022). Consequently, researchers have called for studies that assess more nuanced aspects of nature, such as varying degrees of urbanity and naturalness, proximity to water or topography, and actual use of greenspaces (e.g., Keniger et al., 2013; Hartig et al., 2014; Bratman et al., 2015; de Keijzer et al., 2016). In response to these recommendations, we conceptualized the outdoor environment in terms of the surrounding urbanity and natural amenities. Urbanity reflects the population size and density of an area. Natural amenities are geographic and weather-related characteristics typically considered as desirable, including moderate temperatures, low humidity, topographic variation (e.g., mountains), and bodies of water (McGranahan, 1999).

Further, we explored three aspects of the experience of being outside – amount of time spent outside, enjoyment of time spent outside, and outdoor activities. We used a concurrent mixed-methods approach, in which quantitative data (i.e., objective geographic data and survey responses) and qualitative data (i.e., responses to open-ended survey questions) were collected at the same time, and each used in the interpretation of the results (Clark and Creswell, 2008; Pluye and Hong, 2014). Our primary goal of the study was to understand how aspects of the built and natural environment, and outdoor experiences, are associated with employee work outcomes. To answer this question, quantitative data were used to demonstrate features of the built and natural environment, elucidate how much time employees spend outside during work and nonwork time, and identify the extent to which they enjoy the time they spend outside. Analyses of the qualitative data were used to classify the specific activities in which individuals engaged while they were outside. Findings from the qualitative responses therefore complement the primary research question assessed via quantitative surveys by providing more nuanced information about how participants actually spent time outside, thus capturing a unique and important piece of the outdoor experience. This qualitative approach contrasts with existing work on outdoor activities which typically uses researcher-generated categories in a checklist format, which may not include all types of activities individuals engage in outside. Instead, asking participants how they spent their time outside yields more specific and accurate information that is less influenced by potential researcher bias (i.e., assumptions about how individuals spend their time outside) than if pre-determined categories were used, highlighting an advantage of mixed-methods approaches (e.g., Mazzola et al., 2011). Overall, in combination with the quantitative model, the qualitative data were used to identify the various outdoor activities individuals performed at work and at home, whether activities were comparable or different in these distinct domains, and how activities were associated with work outcomes.

Beyond the amount of time spent outside and the enjoyment of time spent outside, how people spend their time outside is another important feature of their experience outdoors. In past work, researchers have focused on specific types of outdoor activities, typically physical activities, such as exercising outside (Barton and Pretty, 2010) or walking in a park (e.g., Sianoja et al., 2018). Others have explored a wider range of activities individuals engage in outside, but very little work has examined outdoor activities during work time (e.g., Petersson Troije et al., 2021), or combined both work and nonwork time use (e.g., Lottrup et al., 2012; Korpela et al., 2014; Hyvönen et al., 2018). Stigsdotter and Grahn’s (2011) study included a comprehensive list of outdoor activities, but they explored which activities were preferred by Swedish individuals who experienced high levels of stress. Kellert et al. (2017) indicated that walking and hiking are American adults’ favorite outdoor activities, but did not examine the specific outdoor activities in which individuals engaged, nor did they consider activities during both work and nonwork time. Further, despite substantial research on general time-use (e.g., Stinson, 1999), less is known about outdoor-specific activities. Overall, it is unclear how individuals spend time outside during work time, and how this compares to their nonwork time spent outside.

Certain types of outdoor activities may be associated with favorable work outcomes. Oppezzo and Schwartz’s (2014) four-part study found that walking has a positive effect on general creative thinking, and that walking outside, compared to indoor environments, is particularly beneficial for idea generation. Other research suggests that walking outside can improve concentration at work, as well as general cognitive performance outcomes and mood (e.g., Berman et al., 2008; Sianoja et al., 2018), factors that can influence job engagement and creativity. White et al. (2013) found that participants who visited nature reported feelings of restoration when walking, and that nature visits with children were less restorative than visiting nature alone. Psychological benefits have also been found for other types of outdoor physical activities (e.g., swimming, running, cycling; Barton and Pretty, 2010; Korpela et al., 2017a). However, most of this past work has focused primarily on physical activities, has not examined outdoor activities at work, and it is unclear whether these benefits translate to work-related outcomes.

One noteworthy exception is Hyvönen et al.’s (2018) person-centered study, which revealed that Finnish employees in profiles characterized by the most frequent visits to natural environments during work and leisure time experienced greater engagement at work than those in profiles characterized by the least frequent visits (Hyvönen et al., 2018). Across all profiles in their study, most participants reported “enjoying nature and natural scenery,” “relaxing and dwelling,” and “walking and jogging,” as leisure outdoor activities (Hyvönen et al., 2018). Due to this, profiles were characterized by the frequency of time spent in natural settings and variability of activities engaged in, rather than the specific type of activities. In work-specific studies, researchers have explored how general activities (i.e., not outdoor-specific) relate to energy replenishment. Trougakos et al. (2014) found that relaxing lunch breaks reduced fatigue at the end of the workday, whereas social and work-related breaks increased fatigue, particularly for employees with low autonomy. In Bennett et al.’s (2020) study, employees who engaged in microbreaks at work that allowed for mental disengagement from one’s work (i.e., detachment) reported reduced fatigue. Other work has found that spending time on leisure activities during nonwork time (e.g., social, low effort, or physical activities), but not work or household related activities, predicts greater engagement at work (ten Brummelhuis and Bakker, 2012). Collectively, previous research has not explored outdoor activities, both at work and home, alongside work-specific outcomes. It is therefore unknown whether specific activities are associated with engagement and creativity at work.

We collected three sets of data for this study. An employee sample was collected from Amazon’s Mechanical Turk (MTurk; December 2019–January 2020, N = 564), and two samples were collected from employed undergraduate students (N = 239), one in the fall semester (December 2019; N = 90), and one in the following spring semester (March 2020–May 2020; N = 149). Participants recruited from MTurk lived and worked in different geographic regions across the United States, whereas working student participants were recruited from a large state university in northern Colorado (see Figure 2 for heat map of participant locations). All participants completed the same 15-min online survey and responses from all three data collections were combined for analyses (N = 803). The samples were combined because (a) all students were enrolled at the same university, so the built and natural environment where students work and live was homogeneous, thus limiting the variance in urbanity and natural amenities, (b) combining samples produces a larger sample size and greater power to detect indirect effects (Fritz and MacKinnon, 2007), and (c) to account for potential season (i.e., winter versus spring) and COVID-related effects. As explained below, demographic characteristics that vary between samples were modeled as control variables, thus accounting statistically for any differences in demographic characteristics between samples (e.g., age, work hours). Although we combined the samples, we also conducted analyses among the separate samples to investigate whether the decision to combine samples affected the results.

First, participants who failed either of the two attention checks included in the survey (e.g., “select strongly disagree”) were excluded. Participants who reported working less than 10 h last week were also excluded, given our interest in work-related outcomes. Next, missing data were explored and addressed following Newman’s (2014) procedures. Univariate outliers were identified and removed using procedures outlined in Tabachnick and Fidell (2013). We report and interpret the results with outliers excluded, given that they reflect participant error (e.g., careless responding or misunderstanding), or participants who were not representative of the intended sample (e.g., individuals who spend large amounts of time outside are considerably different than American adults’ typical outdoor time use; Klepeis et al., 2001; Kellert et al., 2017).

First, to assess the internal structure and psychometric properties of the study’s latent variables, confirmatory factor analyses (CFAs) were performed for the job engagement scale and the creativity at work scale. Recommendations from Hu and Bentler (1999) and Yu (2002) were used to assess model fit. Specifically, model fit values near the following cutoffs indicate good model fit: a non-significant χ² statistic (i.e., value of p greater than 0.05), CFI greater than or equal to 0.95, TLI greater than or equal to 0.95, RMSEA less than or equal to 0.06, and SRMR less than or equal to 0.08.

Fully saturated mediated regression path models in Mplus Version 8 were used to test the hypothesized indirect effects of urbanity and natural amenities at work and home on job engagement and creativity, via amount and enjoyment of time spent outside in the last week, accounting for controls. Bootstrapping with 5,000 bias-corrected bootstrapped samples was employed, based on Preacher and Hayes (2008) recommendations, and given that this method is the most powerful for tests of mediation (Fritz and MacKinnon, 2007). Asymmetrical confidence intervals were assessed because they more accurately represent the distribution of the product of coefficients, which is not normally distributed (MacKinnon et al., 2004). Therefore, significance of indirect effects was determined by 95% bias-corrected bootstrapped asymmetrical confidence intervals that did not include zero. See Table 1 for sample sizes, descriptive statistics, and correlations among study variables.

For the qualitative responses, content analysis was used to identify outdoor activity categories (Schreier, 2012). Content analysis is particularly useful when analyzing large amounts of text-based data and when the unit of analysis is a simple word or phrase (e.g., Krippendorff, 2004; Braun and Clarke, 2006). In line with best practices for qualitative coding, a series of precoding, coding, and recoding activities were performed in a collaborative process conducted by two independent coders (e.g., Ritchie et al., 2013; Saldaña, 2016; Ravitch and Carl, 2019). Our detailed codebook included mutually exclusive categories with descriptive names, definitions, inclusion and exclusion criteria, typical examples, and “close, but no” examples (Bernard and Ryan, 2010; Schreier, 2012). The final round of coding was conducted for all 803 participants, with 24,090 coding decisions, resulting in 98% overall agreement and an overall kappa value of 0.92, which is considered a nearly perfect level of agreement (McHugh, 2012). Associations between outdoor activities and work outcomes, with all control variables included, were explored using linear regression analyses in SPSS Version 26.

A single-factor CFA for the job engagement scale had poor model fit; χ²(135) = 4101.42, p < 0.05, CFI = 0.67, TLI = 0.63, RMSEA = 0.19, SRMR = 0.11. In line with the theoretical conceptualization of job engagement as having dimensions for emotional, cognitive, and physical engagement at work, a three-factor model was tested next. The three-factor CFA had adequate model fit; χ²(132) = 1017.14, p < 0.05, CFI = 0.93, TLI = 0.92, RMSEA = 0.09, SRMR = 0.04, and demonstrated significant improvement in model fit compared to the single-factor model (Δ χ² = 3084.28, p < 0.001). Therefore, three separate dimensions of job engagement (emotional, cognitive, and physical) are examined across analyses. The single-factor CFA for the creativity at work measure had sufficient model fit; χ²(65) = 515.94, p < 0.05, CFI = 0.94, TLI = 0.93, RMSEA = 0.09, SRMR = 0.03.

Tables 2, 3 depict the direct effects among the built and natural environment (i.e., urbanity and natural amenities), mediators (i.e., time outside and enjoyment of time outside, respectively), and work-related outcomes. Table 4 provides the 95% confidence intervals (CIs) for the indirect effects; CIs that do not include zero (bolded in the table) reflect significant indirect effects.

Although findings may be reported entirely qualitatively (e.g., describing themes and categories), quantifying qualitative data is common in content analysis and can be effective for studies with large sample sizes (Krippendorff, 2004; Braun and Clarke, 2006; Schreier, 2012; Pluye and Hong, 2014). Therefore, we report descriptive quantitative findings gleaned from the qualitative responses (see Figures 3, 4), and test associations between outdoor activities and work-related outcomes.

To explore whether different types of outdoor activities were associated with the environment where participants live and work, or their experience of the environment, supplemental point-biserial correlational analyses were performed (Tabachnick and Fidell, 2013) and presented in Table 5.

There are a number of ways in which the results from our study can inform practical recommendations for organizations, supervisors, and employees who are interested in improving job engagement and creativity at work. Yet, it is important to note that the magnitude of the significant effects detected in the present study were small, suggesting that the following practical recommendations may be best interpreted provisionally.

At the employee-level, our results suggest that individuals may experience greater engagement and creativity at work if they spend more time outside at work and at home and enjoy the time they spend outside. Certain types of outdoor activities may predict greater time spent outside (e.g., exercising during nonwork time), enjoyment of time outside (e.g., relaxing activities), and job engagement (e.g., social activities during nonwork time). These individual-level approaches would be cost-effective for organizations and could be further reinforced by supervisors who can effectively help employees balance the demands of their work and nonwork lives (Hammer et al., 2009) and encourage employees to spend time outside (e.g., by having outdoor meetings).

At the organizational-level, organizations can implement nature-related policies, such as paid days off to spend outside or discounted State and National Park passes. Instituting and enforcing flexible schedule policies can also enable employees to spend time outside and in enjoyable ways; flexible schedules, including longer breaks during work hours (while not reducing paid work time), could allow employees who want to surf in the morning, hike in the afternoon, sunbathe mid-day, or stargaze in the evening to do so. More substantial workplace redesign efforts can include creating rooftop gardens, installing on-site bike racks, picnic tables, and outdoor game spaces, or creating nearby paved walking or biking trails with wheelchair accessibility. It is important that on-site outdoor spaces are close in proximity to the workplace, comfortable, and suitable for different weather conditions (e.g., include fans in hotter climates, ensure different forms of sun protection) (Petersson Troije et al., 2021).

The primary limitations of this study are methodological, which can be addressed in future research. First, a cross-sectional design was used in this formative study to establish whether outdoor environments and experiences are associated with work outcomes, so an important next step would be to explore similar research questions using more advanced methods, such as longitudinal designs or experience sampling methodology (ESM). For example, a longitudinal study could assess the different places individuals work and live across their life course and examine how different locations influence experiences of the outdoors and work-related variables. In an ESM study, participants’ real-time locations could be reported or assessed using geographic data, while simultaneously assessing length of time outside, current levels of enjoyment, and activities engaged in, and then be examined alongside same- or next-day work outcomes. As previously mentioned, the effects detected in the present study are small. However, given the multitude of variables that can influence an employee’s engagement and creativity at work, it is not surprising that we found small effects related to the built and natural environment and experiences outside on these outcomes. Moreover, “moving the needle” even marginally on work-related outcomes is meaningful for both employees and organizations, particularly for modifications that can be low-cost and low-effort, like going outside. Conducting more rigorous studies using advanced designs may yield more precise effects that are stronger in magnitude.

In addition, the use of objective location data is a strength of the study because it is more reliable and unbiased than a participant’s perception of the degree of urbanity or natural amenities where they live and work. However, other variables were self-reported by participants (e.g., time spent outside, enjoyment of time outside, job engagement, creativity), introducing the possibility for common method bias (Podsakoff et al., 2012). Future research could employ the use of supervisor reports or customer satisfaction ratings as indicators of employees’ engagement and creativity (e.g., Wildman et al., 2011; Mumford and Todd, 2019). Another next step would be for researchers to further extend the conceptualization of contact with nature by exploring other features of the outdoor environment (e.g., tree canopy coverage, biodiversity; Keniger et al., 2013). Prior research has demonstrated the unique value of exposure to blue spaces (i.e., natural or human-made water features; lakes, ponds, fountains) as a promotive tool for improving health (Smith et al., 2021). Notably, exposure to combined blue-green spaces is a particularly beneficial coping tool during times of societal distress (e.g., COVID-19; Pouso et al., 2021), which is critical given over a quarter of Americans state that they are so stressed they cannot function, often citing societal stressors such as inflation, violence and crime, and political and racial climates (American Psychological Association [APA], 2022). Therefore, it may be worthwhile to explore nuances related to the degree of “greenness” and “blueness” of natural areas.

Additionally, the level of contact with nature at work is likely related to individuals’ jobs and industries, as jobs can vary according to the level of contact with nature from the job context and job tasks (Klotz and Bolino, 2021). For instance, warehouse forklift operators have low contact with nature at work in both their job context and job tasks, florists have high contact with nature in terms of job tasks but not job context, and farmers have high contact with nature in their job tasks and context (Klotz and Bolino, 2021). It may also be worthwhile to examine individuals whose job provides high levels of contact with green spaces (e.g., landscapers, botanists) compared to blue spaces (e.g., commercial divers, marine biologists). Future research could focus on a specific occupation or industry or examine job type as a substantive variable that may influence nature contact, the experience of time spent outside, and work outcomes.

Another potential methodological drawback is the combination of working students and participants collected on MTurk in the analyses. As previously noted, this decision was due to the students being co-located (thereby limiting the variance in the location-related predictor variables of urbanity and natural amenities), the need for a large enough sample size to have adequate power to detect indirect effects (Fritz and MacKinnon, 2007), and to account for potential season and COVID-related effects, as the data collections spanned the fall, winter, and spring months, and reflected the time period both before and during the COVID-19 pandemic (81.5% of the participants completed the survey before COVID). Although all participants reflected working adults, there were some differences between samples. Unsurprisingly, compared to participants in the MTurk sample, working students were younger, worked fewer hours, and were less likely to work a regular daytime schedule. Accordingly, we statistically accounted for the different samples as well as relevant characteristics (e.g., age). Additionally, we ran the mediation analyses with only the MTurk sample and found nearly identical results. In line with our previous recommendations, future work would benefit from examining employees working in the same type of job, rather than the broad approach we took in this study. Additionally, although not a focus of this study, it would also be interesting for future work to examine potential differences in outdoor experiences and related job outcomes prior to, compared to during, crises such as pandemics.

Alternative mechanisms that explain the associations found in the present study could be explored in future work. For example, it is possible that positive affect or emotions are acting as additional mediators, as nature can restore emotional energies (Klotz and Bolino, 2021), and having more leisure time and enjoying oneself can produce positive feelings (Newman et al., 2014), which can relate to greater engagement and creativity at work (Hennessey and Amabile, 2010; Bledow et al., 2011). Another possible explanatory mechanism is sleep, as sleep can be influenced by nature and time spent outside (Shin et al., 2020). American adults who live in areas with greater natural amenities and greenspace are more likely to report sufficient sleep (Grigsby-Toussaint et al., 2015) and have lower odds of short sleep (Johnson et al., 2018). Similar to contact with nature, healthy sleep is also critical for energy replenishment and related emotional, cognitive, and physical outcomes (e.g., Crain et al., 2018). Future research should also consider other aspects of workers’ health and well-being. For example, how does spending time outside during work and nonwork hours relate to job attitudes and the physical and mental health of employees? Is spending time outside or engaged in specific outdoor activities beneficial for reducing the deleterious effects of work stressors that are associated with poor mental health, or among workers with mental health conditions (e.g., ADHD and anxiety) (Follmer and Jones, 2018)?

As more research is being conducted that investigates issues related to climate change and work (e.g., Kjellstrom et al., 2016), understanding the interaction between work and engaging with nature is ripe for additional studies. For example, how might results vary based on climate in geographical regions that differ in terms of temperature, humidity, rate and types of natural disasters, or hours of sunlight? Are there differences across occupations that vary in the extent to which aspects of the work itself may be performed outside, and how might climate, and climate changes, relate to those experiences?

Researchers should consider other work outcomes in future studies. Klotz and Bolino (2021) theorize that contact with nature at work should also increase prosocial energy, which allows individuals to invest resources in the well-being of others. This is in line with the reasonable person model, which stipulates how supportive environments (which are often natural) enable reasonableness, cooperation, and helpfulness (Kaplan and Basu, 2015). These ideas about how nature can influence prosocial behaviors have gained some empirical support (e.g., Zhang et al., 2014; Zelenski et al., 2015). Individuals who have greater contact with nature at work and home may also exhibit more prosocial behaviors at work and fewer deviant behaviors at work. Examining other work-related variables is a promising avenue for future research.