Do Preferred Landscapes Reduce Stress?

Abstract

Humans may prefer landscapes that promote health (Kaplan, 1998). We know that humans feel less stress¬ful in green environments (Jiang, 2014; Hartig et al., 2014; Hartig, 2003; Simons, 1991). We also know that humans like green environment (Nassauer, 1995; Lovell & Sullivan, 2006; Gobster, Nassauer, Daniel, & Fry, 2007), but the connection between landscape prefer¬ence and stress reduction has not been fully established. If such a link were established, landscape architects, planners, and land managers would have a useful, fast tool for understanding the impact of urban designs on a city’s inhabitants. In this study, we explore the extent to which preference for a landscape predicts the capac¬ity of people to recover from a stressful experience. The study involves a preference questionnaire and a lab experiment. First, participants completed a preference questionnaire showing neighborhood street scenes in the U.S. Midwest. Then the same participants experienced a stressful event and then watched one of ten 3-D videos of neighborhood street scenes with varying levels of tree cover. We measured changes in participants’ stress levels through three response pathways: Hormonal (salivary cortisol), physiological (skin conductance) and psychological (self-reported stress). The study has been conducted with consent, and there is no lasting psy¬chological effect to the participants. We examined the extent to which preference predicted stress reduction in 160 participants through linear regressions. When both genders were analyzed together, there were associations between preference and stress reduction including aver¬age stress recovery [F(1,158)=5.419, R2= .03, p< .05 ] and psychological pathway (VAS) [F(1, 158)=6.559, R2= .04, p< .05]. Separate analysis for men and women showed a clear gender difference. For men, we found a significant positive association between preference and physical response [F (1, 78) = 4.833, R2= .10, p< .05] but no association for hormonal or psychological responses. For women, we found a significant posi¬tive association between preference and psychological responses [F (1, 78) = 4.823, R2= .05, p< .05] but no association for physiological or hormonal responses. These findings are consistent with other results showing that men and women recover from stressful events in different ways and along different timelines. They also suggest that more highly preferred settings are associ¬ated with greater recoveries from stress, but through different pathways for men and women.