Enhancing wellbeing with psychological tasks along forest trails

1 1 Running head: PSYCHOLOGICAL TASKS ALONG FOREST TRAILS 2 Enhancing wellbeing with psychological tasks along forest trails 3 1Kalevi KORPELA, 2Eira-Maija SAVONEN, 1Suvi ANTTILA, 4 1Tytti PASANEN, 1Eleanor RATCLIFFE 5 Citation: 6 Korpela, K. M., Savonen, E.-M., Anttila, S., Pasanen, T., & Ratcliffe, E. (2017). Enhancing wellbeing 7 with psychological tasks along forest trails. Urban Forestry & Urban Greening, 26, 25-30. 8 http://dx.doi.org/10.1016/j.ufug.2017.06.004 9 10 Acknowledgements 11 The establishment of the forest trails and data collection was funded by the European Agricultural 12 Fund for Rural Development and the Council of Tampere Region (in Finland). We are grateful for 13 the volunteers who participated in the study. 14 15 16 Kalevi Korpela (corresponding author), Professor of Psychology, Dr. Psychol.; 17 [email protected]; tel: + 358-50-3186 130 18 1 Faculty of Social Sciences / Psychology, 33014 University of Tampere, Tampere, Finland; 19 2 Natural Resources Institute Finland (LUKE) / 39700 Parkano 20 21 E-mails: eira-maija.sa[email protected]; [email protected]; ty[email protected]; 22 [email protected] 23 24 25 26 27 28 29 2 30 31 Highlights 32 33  We investigated the role of awareness-enhancing tasks in nature experiences. 34  We found no country differences in the restorative experiences after forest walks. 35  Satisfaction with the tasks was linked to restorative change and mood enhancement. 36  Nature-connectedness did not moderate either outcome. 37 38 Abstract 39 40 The purpose of the study was to investigate whether deliberate psychological tasks, intended 41 to focus people’s attention on the interaction between themselves and natural surroundings, are 42 linked with mood enhancement and self-reported restoration. In four European countries (Finland, 43 France, Luxembourg, Sweden), we surveyed the experiences of volunteers (N = 299) who walked 44 forest trails and carried out psychological tasks printed on the signposts along them. We 45 investigated the similarities and differences of the trail experiences between the countries. Via 46 multigroup modeling, we further examined the moderating role of nature-connectedness in 47 relationships between satisfaction with the contents of the psychological tasks, mood enhancement, 48 and restorative benefits. The results showed that, independent of age and gender, participants were 49 more satisfied with the trails in Sweden and Luxembourg than in Finland. We detected no reliable 50 differences in the restorative experiences or willingness to recommend the trail for others. In the 51 moderation model, satisfaction with the signposts’ contents was connected to positive restorative 52 change and mood enhancement. The moderator effects of nature-connectedness were not significant 53 for either outcome. Thus, it is likely that satisfactory tasks will work equally well for people varying 54 in nature-connectedness. This is a promising prospect for public health promotion. The fairly high 55 level of nature-connectedness among the participants limits the generalizability of our results. 56 Conclusions concerning the role of nature-connectedness should be made with caution due to the 57 limited coverage of the concept in our measure. Future studies that separate the effect of 58 psychological tasks from the restorative effects of nature itself are needed. 59 60 Keywords: engagement; psychological tasks; forest trail; mood; restoration; nature-connectedness 61 3 62 63 64 65 Introduction 66 67 Links between nature and wellbeing, and particularly benefits to mood, stress reduction and 68 attention restoration, are well-explored in the literature within environmental psychology (see 69 Hartig, Mitchell, de Vries, & Frumkin, 2014, for a review). However, it is not reliably known 70 whether deliberate tasks intended to focus people’s attention to the interaction between themselves 71 and natural surroundings can enhance mood and restoration from stress. Such an engagement-based 72 approach has been used, however, in two studies (Duvall, 2011, 2013). To test whether conscious 73 engagement strategies enhance the benefits of nature walks, Duvall (2011) conducted a two-week 74 walking experiment with two treatment conditions: “schedule setting” and “awareness plans”. All 75 participants were given the goal of taking at least three 30-minute outdoor walks during each of the 76 subsequent two weeks. The schedule setting group identified the days and times they intended to 77 walk whereas those who were assigned to the engagement group were instructed to focus on 78 awareness plans during their walks. These awareness plans instructed the participants to focus on 79 the surrounding physical environment in a variety of ways, such as focusing on their senses or 80 taking on a new role (e.g., imagining you are an artist looking for inspiration). The results showed 81 that using awareness-enhancing tasks increased satisfaction and self-reported attentional 82 functioning, and decreased feelings of impatience and irritation. This suggests that directing 83 attention and interaction consciously in everyday nature experiences might promote well-being. 84 To investigate the potential of awareness-enhancing tasks in nature experiences, we undertook 85 an international project together with participants from Finland, France, Luxembourg, and Sweden 86 during 2012-2013 (Savonen & Korpela, 2013). Signposts with psychological guidance for 87 perceiving the surroundings, stress-alleviation, and restoration were put up along the forest trails in 88 the four countries. Our rationale was that if restoration from stress can be enhanced with 89 psychological tasks along forest trails, such interventions could be carried out on a larger scale to 90 benefit public health promotion in an attempt to reduce the accumulation of stress cost-effectively. 91 Scientific investigations about the effectiveness of such trails are nonexistent except for a 92 preliminary user survey of the Finnish trail (Vattulainen et al., 2011). In that survey, 69% of the 93 participants reported that they were calmer, more alert and more away from everyday worries than 94 4 before walking. Seventy-three percent of the visitors were very or quite satisfied with the trail they 95 walked (Vattulainen et al., 2011). 96 In the present study, we survey the experiences of the volunteers who walked the forest trails 97 and carried out the psychological “tasks” along them. At this point, we were not able to set up an 98 experimental design to tease out the effects of completing the tasks versus walking the trails without 99 them. However, by comparing the survey answers from the different countries our first aim is to 100 investigate whether different trails – with different types of forests and landscapes and culturally 101 different participants – produce similar experiences. Similarities in experiences might arise from the 102 fact that all participants carried out the same psychological tasks while walking on the trails. 103 The contents of the tasks were planned according to an attention restoration theory (ART) by 104 the Kaplans (1989) and stress reduction theory (SRT) by Ulrich (1983). Both mood benefits and 105 restorative changes after nature exposure are well-documented in the current literature (Bratman, 106 Hamilton & Daily, 2012; Hartig et al., 2014). A consistent finding in experimental studies is that 107 even passive viewing of urban parks or woodlands (compared to built urban environments) after 108 negative antecedent conditions, such as attentional fatigue (Kaplan & Kaplan, 1989) and 109 psychophysiological stress (Ulrich et al., 1991), produces greater physiological changes toward 110 relaxation, greater changes to positively-toned self-reported emotions, and faster recovery of 111 attention-demanding cognitive performances (Hartig et al., 1991; Parsons, et al., 1998; Tsunetsugu 112 et al., 2013; Tyrväinen et al., 2014; Ulrich et al., 1991). Evidence of improved attentional 113 functioning, emotional gains, lowered blood pressure, and salivary cortisol has also been reported in 114 field experiments using actual walks in urban parks or woodlands and repeated measures of 115 physiological, emotional, and attentional outcomes (Hartig et al., 1991, 2003; Park et al., 2010). 116 In planning the order of the tasks, we paid attention to the current knowledge of the order of 117 appearance of the restorative outcomes. The timeline of the appearance of restorative, non-stressing 118 outcomes after the exposure (watching or walking) to the natural environment (park, forest) follows 119 the physiological and cognitive response patterns reported for stress-related changes: a decrease in 120 heart rate, muscle tension and skin conductance within four to seven minutes (Ulrich et al., 1991), 121 lower blood pressure and improved mood after 20 minutes (Hartig et al., 2003; van den Berg et al., 122 2003), and better attentional performance after 10-55 minutes (Berman et al., 2008; Hartig et al., 123 2003; Faber Taylor & Kuo, 2009). According to ART, a person may pass through successive levels 124 of the restorative experience (Kaplan & Kaplan, 1989). The first level involves ‘clearing the head’, 125 leading to increased directed attention capacity. The second entails facing accumulated matters on 126 one’s mind. The third is reflection on one’s priorities, prospects, actions, and goals in life. 127 5 In addition to restoration theories, we utilized environmental self-regulation theory based on 128 favourite place studies (Korpela, 1989) in developing the tasks. Favourite places are places to which 129 people are emotionally attached and, for example, in Finland, 40% - 48% of adults who live in cities 130 have reported that their everyday favourite or positive places are in natural settings (Korpela et al., 131 2001; Kyttä et al., 2013). Favourite places are sought out for restorative and self-regulatory benefits 132 in the service of coping with everyday stress and personal worries arising from, e.g., social, 133 financial or work-related matters (Korpela et al., 2010). 134 There is evidence that self-reported “connectedness with nature” influences the process 135 whereby nature affects mood and restorative outcomes (Capaldi, Dopko, & Zelenski, 2014). We 136 took this as a second focus of our study. Connectedness with nature (Mayer & McPherson Frantz, 137 2004) or nature relatedness (Nisbet, Zelenski, & Murphy, 2009) can be defined as people’s interest 138 in and desire for nature contact; a trait-like or state-like subjective connection to nature or feeling in 139 community with nature (for a review, see Capaldi et al., 2014). Experimental studies have found 140 that increases in emotional well-being through nature exposure are partially due to increased state-141 like nature connectedness, indicating a mediation effect (Mayer, McPherson Frantz, Bruehlman-142 Senecal, & Dolliver, 2009). Conversely, some studies have found that exposure to nature increases 143 positive affect which mediates the increase in state-like nature connectedness (Nisbet & Zelenski, 144 2011). However, the effect of contact with nature on wellbeing (positive affect and feelings related 145 to elevation) has not been moderated by trait-like nature connectedness in an experimental study 146 (Passmore & Howell, 2014). Thus, there is limited knowledge about whether connectedness with 147 nature is the primary cause or a moderating or a mediating factor in the process through which 148 nature affects mood and restorative outcomes (Bratman et al., 2012). More research on mediation 149 and moderation by nature connectedness has been explicitly called for (Capaldi et al., 2014). Our 150 interest in this study is in the moderating role of nature-connectedness which seems to be non-151 existent according to a previous, experimental study (Passmore & Howell, 2014). We want to know 152 whether nature-connectedness moderates mood enhancement differently from other restorative 153 experiences (relaxation, attentional and cognitive restoration). We want to know whether the 154 potential restorative and mood enhancing effects of the psychological tasks along the trails that are 155 intended for the general public are heavily dependent on nature-connectedness. This might mean 156 that only people with a strong interest in and desire for nature might benefit from such tasks and the 157 trails might not serve the general public as widely as would be desirable. 158 In sum, we aim to investigate the similarities and differences of the trail experiences between 159 the countries. Our second aim is to investigate relationships between nature-connectedness, 160 6 satisfaction with the signpost contents, mood enhancement, and restorative benefits. We investigate 161 whether nature-connectedness moderates the effects of satisfaction with the signpost contents on 162 both mood and restorative change. 163 164 Method 165 Sample and procedure 166 The study was conducted in different years as a survey study in Finland, Sweden, 167 Luxembourg, and France near the well-being trails (Table 1). 168 Table 1 here 169 The first well-being trail was opened in May 2010 in Finland. The trail follows hiking tracks 170 and narrow countryside roads in a forest, along lakeside and in a cultural landscape. It includes a 171 pleasant view from a high cliff, overlooking a lake bay. The well-being trail in Luxemburg was 172 opened in April 2013. The trail goes along the already existing, well-established hiking trails in 173 spruce and beech forests, and leads to a small village road opening onto a view of nearby fields. The 174 Swedish well-being trail was opened in September 2012. It is located in spruce and birch forests, 175 where it follows a skiing track used for walking only in a limited period outside the skiing season. 176 The French trail was established into the variable coniferous and deciduous forests in September 177 2013. The trail follows narrow countryside roads and forest paths, and has pleasant, open views 178 over agricultural land. 179 The questionnaires were available at the nearby camping/community/spa reception desks or, 180 in Finland, in the boxes in the beginning and end of the trail. In addition to the variables reported in 181 this paper, the questionnaire included open-ended questions about place of residence, reasons to 182 visit the trail, a signpost remembered well, and situations where the tasks might suit a person best. 183 The participants filled in the questionnaire immediately after the walk. A total of 299 people 184 participated in the study. The number of participants in the four countries were 164 in Finland, 54 in 185 Sweden, 65 in Luxembourg and 16 in France; in all four countries, more females than males 186 completed the survey (Table 2). 187 Table 2 here 188 Well-being trails in each of the countries contained the same nine signposts, set up at intervals 189 and referring to rehearsal tasks to be undertaken along the trail. The first and the last rehearsals 190 7 contained self-rated items about restorative experiences (see Measures). Rehearsal two dealt with 191 relaxing and letting the environment fascinate the walker. The following is a short example of the 192 beginning of the instructions for the relaxation task: “Breathe slowly and let your shoulders relax. 193 Take a look around you and let your mind be attracted by some pleasant spots or details on the 194 ground, in the woods or in the sky…”. At signpost three, the walker was encouraged to observe 195 nature and one’s mood and feel it improve. An extract from the mood instruction is: “Let the 196 scenery affect your mood. Listen to the sounds of nature or the silence and let your mind drift away 197 from the everyday cares/hassles…”. Rehearsal four dealt with finding a quiet place in the 198 surroundings. Rehearsals five and six were related to favourite places. The walker was urged to find 199 a favourite place and then to share it with others or, if alone, to try to form a clear mental image of 200 the place to be remembered in the future. Rehearsal seven dealt with observing and recognizing 201 current mood. At rehearsal signpost eight the walker was urged to observe surroundings and find a 202 metaphor that captured their current life situation. To enhance the potential wellbeing effects, the 203 language used on all of psychological signposts was suggestive and in five signposts contained 204 phrases such as “Feel your mind and body becoming calm” and “Feel your mood improve” 205 (Sinclair, Soldat & Ryan, 1997; Wilson & French, 2014; Wiseman & Greening, 2005). 206 Measures 207 Restorative change. Restorative change was measured with four items tapping relaxation, 208 energy, and attentional and cognitive restoration in accordance with the previous measures and 209 findings of restorative outcomes (Hartig, Lindblom, & Ovefelt, 1998; Kaplan, Bardwell & Slakter, 210 1993; Korpela, Ylén, Tyrväinen & Silvennoinen, 2008; Staats, Kieviet, & Hartig, 2003). The items 211 (“I'm feeling calm and relaxed”, “I'm alert and focused”, “I'm enthusiastic and energetic”, “All my 212 everyday concerns and worries are away”) were printed on the signposts at the start and in the end 213 of the trail. The participants were asked to rate their restorative experiences on five-point Likert 214 scales (1 = not at all, 2 = a little, 3 = quite much, 4 = very much, 5 = fully) and to calculate the sum 215 of their assessment and to write it down. At the end of the trail they were asked to evaluate their 216 experience again with the same items and procedure and report the extraction of these two sums 217 (end – beginning). The extraction represents the change in the restorative experience, negative 218 values representing worsening and positive values representing improvement of restoration. 219 Mood enhancement. Mood enhancement was asked with a single item (“To what degree was 220 your mood enhanced after walking the well-being trail?”) with a 5-point response scale (5 = very 221 much, 4 = rather/quite much, 3 = moderately, 2 = not very much, 1 = not at all). 222 8 Satisfaction with the contents, location and number of the signposts and with the trail overall. 223 Satisfaction was asked with single items (“How satisfied are you with the contents / location / 224 number of the psychological rehearsals along the trail / with the trail in all?”). Satisfaction with 225 contents and location was assessed with 5-point response scales (5 = very satisfied, 4 = quite 226 satisfied; 3 = moderately satisfied; 2 = not very satisfied, 1 = unsatisfied). Satisfaction with the 227 number of signposts was assessed with a 3-point scale (1= I would prefer fewer signposts, 2 = this 228 was a suitable number of signposts, 3 = I would prefer more signposts). 229 Willingness to revisit and recommend to friends. These were asked with single items (“How 230 willingly would you visit and walk the trail again?”, “Would you recommend the trail to your 231 friends and acquaintances?”) and responded with 5-point scales (1 = not at all, 2 = not very 232 willingly, 3 = moderately willingly, 4 = quite willingly, 5 = very willingly). 233 Nature connectedness. As our questionnaire in field conditions had to be brief we developed a 234 nature connectedness scale for the purposes of this study in line with earlier studies on place 235 attachment, connectedness or relatedness with nature (cf. Kyle, Mowen, & Tarrant, 2004; Mayer & 236 McPherson Frantz, 2004; Nisbet et al., 2009; Tyrväinen et al., 2007), and urban-related identity 237 (Lalli, 1992). Nature connectedness was measured with three items selected for the present study (“I 238 prefer to be physically active outdoors in nature“, “I sometimes feel a compelling urge to get to 239 nature”, “I prefer to spend my free time in parks and green spaces rather than in the urban (built) 240 environment”) and were rated on a five-point Likert scale (1=Totally disagree, 2=Disagree quite 241 much, 3=Neither disagree nor agree, 4=Agree quite much, 5=Totally agree; Cronbach’s alpha = .80 242 in Finland, .71 in Sweden, and .55 in Luxembourg). We used the mean summary score of the scale 243 in ANOVAs. 244 Age and gender. Age (“My age is...”) was asked open-endedly and gender with two options 245 (male/female). 246 247 Statistical analyses 248 One-way ANOVAs and ANCOVAs were used for comparisons between Finland, Sweden and 249 Luxembourg. France was excluded from these analyses due to small sample size. We calculated the 250 sample size with Gpower (version 3.1.9.2, Faul, Erdfelder, Lang, & Buchner, 2007) using the effect 251 size of .31 reported by McMahan & Estes (2015) for positive affect after nature exposure. For one-252 way ANCOVA with three countries and two covariates, the required sample size for detecting 253 9 group differences at a .05 risk level and with .95 power is 164. For post hoc comparisons, we used 254 Bonferroni test and with unequal variances Dunnett’s T3 test. 255 Moderation analyses were conducted using Mplus version 7 with WLSMV estimator that 256 allows the estimation of ordered categorical outcomes. To examine country-level differences, we 257 first tested multigroup models with all paths separately estimated in each country. Then, the paths 258 were constrained to equal across countries, one at a time, until all paths were the same in each 259 country and a single model instead of a multigroup model was specified. The moderation model that 260 had the best overall fit with the data (based on χ2 difference tests) was interpreted. However, the 261 first (with three separate moderations in each country) and the final (single) models were saturated, 262 that is, their degrees of freedom was 0, and no fit indices could be calculated for them (Bollen, 263 1989). The other models were considered acceptable if they satisfied the following criteria: the 264 χ2test (Bollen, 1989), Comparative Fit Index (CFI) with values greater than .95, Root Mean Square 265 Error of Approximation (RMSEA) with values smaller than .05, and the absolute values of the 266 correlation residuals smaller than .10 (Kline, 2016). 267 For the analyses, two cases were deleted due to extreme restorative changes, from not feeling 268 restored at all by any of the indicators to feeling completely restored by all the indicators (or vice 269 versa) which we found implausible. Mood enhancement, being measured with a 5-point scale, was 270 specified as an ordered categorical outcome in moderation analyses. In moderation analyses, nature-271 connectedness and satisfaction with the signposts’ contents were centered so that their interaction 272 could be examined. In the final model, we interpreted estimates and their 95% bootstrapped 273 confidence intervals. 274 275 Results 276 277 The means of the outcome variables were compared between Sweden, Luxembourg and 278 Finland (Table 3). ANOVA revealed significant F-test values in the change in restorative 279 experiences, overall satisfaction with the trail, willingness to recommend the trail to friends, and 280 satisfaction with the number of the signposts. Although the overall ANOVA was significant in the 281 restorative change and willingness to recommend the trail to friends, according to the Dunnett’s test 282 (Bonferroni in the latter) pairwise comparisons between the countries did not reach statistical 283 significance. However, participants from Finland rated the overall satisfaction lower than Sweden 284 16 Wiseman, R., & Greening, E. (2005). It’s still bending: Verbal suggestion and alleged 482 psychokinetic ability. British Journal of Psychology, 96, 115–127. 483 http://doi.org/10.1348/00071 2604X15428 484 485 486 487 488 489 490 491 492 493 494 495 FIGURE LEGENDS 496 497 Figure 1. Moderation model with WLSMV estimator; standardised estimates and their 95% 498 bootstrapped confidence intervals. Solid lines: estimate differs from 0. 499 500 501 502 17 Figure 1. 503 504 505 506 507 508 18 Table 1. Characteristics of the forest trails in the four countries. 509 Country Area Location Trail length (km) Data collection period Participants Finland Pirkanmaa Ikaalinen spa 4.4 or 6.6 Aug 2010 – Oct 2010 164 Sweden Sunne, Värmland Selma spa 5.0 Oct 2012 – Nov 2013 54 Luxembourg Müllerthal region Nommern 4.3 May 2013 – April 2014 65 France Brouvelieures Frémifontaine 4.7 Nov 2013 – June 2014 16 510 19 Table 2. The frequencies and percentages of the respondents’ gender and the means, range, standard deviations and medians of the trail visitors’ age in years by country. Female Male Total Age, yrs n % n % N M Range SD Md Sweden 46 85.2 8 14.8 54 45.4 17-73 15.0 48.5 Luxembourg 44 67.7 21 32.3 65 48.0 13-82 14.1 49.0 France 10 62.5 6 37.5 16 59.3 23-80 17.8 65.5 Finland 120 73.2 44 26.8 164 47.4 13-66 11.1 50.0 Total 220 79 299 20 Table 3. Analyses of variance (ANOVA) and post hoc pairwise comparisons in three countries (N = 283; France excluded). Country ANOVA Pairwise comparisons Sweden Luxembourg Finland Scales 1-5: M SD M SD M SD F df p Restorative change 2.42 2.07 3.02 3.41 1.76 2.20 4.47 2, 206 .042 n.s. D Mood enhancement 3.25 .96 3.32 .96 3.12 .85 1.26 2, 274 .287 Overall satisfaction with the trail 4.21 .95 4.13 .85 3.80 .75 7.05 2, 274 .001 S>F .005 L>F .02 B B Willingness to revisit 3.62 1.06 3.92 1.15 3.69 1.01 1.46 2, 278 .233 Willingness to recommend 4.09 .92 3.92 1.09 3.72 1.01 3.08 2, 279 .047 n.s. B Satisfaction with the contents of the psychological signposts 3.53 .85 3.63 .83 3.45 .76 1.09 2, 277 .338 Satisfaction with the number of the psychological signposts 1.83 .545 2.08 .378 2.14 .398 10.67 2, 275 <.001 F>S .001 L>S .017 D D Nature connectedness (1-5) 4.16 .74 4.19 .67 4.31 .76 1.18 2, 280 .310 Note. D = Dunnett’s T3 test for unequal variances, B = Bonferroni test 21 Table 4. Spearman correlations (rho) of the study variables (n = 214-295). 1 2 3 4 5 6 7 1. Restorative change 2. Mood enhancement .42** 3. Overall satisfaction with the trail .29** .47** 4. Willingness to revisit .29** .57** .63** 5. Willingness to recommend .31** .47** .64** .76** 6. Satisfaction with the contents of the signposts .24** .32** .52** .50** .49** 7. Satisfaction with the number of the signposts -.06 -.05 -.04 .002 -.03 .08 8. Nature connectedness .14* .22** .16** .20** .24** .13* -.03 Note. * p < .05 , ** p < .01