Where to recover from attentional fatigue: An expectancy-value analysis of environmental preference

Abstract

Preferences for natural and urban environments can be framed in terms of (1) beliefs about the likelihood of psychological restoration during a walk in each type of environment and (2) the evaluation of restoration given differing restoration needs. We conducted an experiment to test hypotheses about restoration as a basis for environmental preferences. Imagining themselves as attentionally fatigued or fully refreshed, participants (N=101) evaluated recovery, reflection, and social stimulation outcomes. Next, they viewed slides simulating a walk through a forest or an urban center, then rated the likelihood of recovery, reflection, and social stimulation outcomes following such a walk. This procedure was repeated with the second environment. Preference for the forest over the city was twice as strong given attentional fatigue. The greater likelihood of restoration in the natural environment in conjunction with more positive evaluation of recovery when fatigued appears to explain this pattern. The results have implications for environmental preference conceptualizations and our understanding of the relationship between preference and restoration.

Introduction

European and North American adult samples consistently prefer unthreatening natural environments over commonplace urban environments (Knopf, 1987; Ulrich, 1983). Whether built primarily on cultural or evolutionary assumptions, explanations for this difference in preferences commonly refer to the relative capacity of natural and urban environments to support consistently positive (e.g. aesthetic, recreational) experiences and so psychological well-being more generally (e.g. Kaplan & Kaplan, 1989). Going beyond assumptions about preference as a marker of salutary quality, some theories explicitly address aspects of psychological well-being that are affected by natural and urban environments (Hartig & Evans, 1993). More specifically, these theories refer to undesirable, potentially harmful antecedent conditions that are better remedied in natural environments. So, although they variously emphasize stress reduction (Ulrich, 1983; Ulrich, Simons, Losito, Fiorito, Miles, & Zelson, 1991; see also Wohlwill, 1983) and recovery from attentional fatigue (Kaplan & Kaplan, 1989; Kaplan, 1995), these theories emphasize the relative standing of natural environments as aids in recovery from undesirable states in which functional capabilities are compromised. They are, in other words, theories about restoration.

Our exploration of the link between preference and restoration starts from attention restoration theory (ART; Kaplan, 1995; see also Kaplan & Kaplan, 1989). ART proposes that after a period of prolonged use of directed attention, a person's capacity to ward off distractions becomes exhausted. An attentionally fatigued person has difficulty concentrating, suffers from increased irritability, and is prone to errors on cognitive tasks. He or she would then benefit from a stay in an environment that does not require reliance on directed attention. When there, the person can rest the inhibitory mechanism on which directed attention depends and so recover the capacity to direct attention. As described in ART, restoration proceeds to the degree that four qualities characterize the person–environment exchange. First, the person has a sense of being away. Psychologically, if not geographically, the person has distance from unwanted distractions and routines that impose demands for directed attention. Second, the person's attention is engaged, effortlessly and to some extent involuntarily, by interesting environmental contents and processes like exploration. As it provides a basis for functioning that does not require inhibitory effort, fascination enables rest of the inhibitory mechanism. It is thought to be prolonged when the environment allows a sense of extent, itself attributed to coherence in the experience of the environment (e.g. the environment allows structure and order in perception, and so is not confusing) and the scope for continued exploration. As a motivational context in which being away, fascination, and extent can occur, compatibility refers to the correspondence between what the person wants to do, must do, and can do in the environment.

Although the proposed restorative qualities can in theory be perceived in any environment, Kaplan and Kaplan (1989) state reasons why they should be more characteristic of natural environments than other types of environments. There is also some empirical evidence that they are perceived to a greater degree in various natural environments, in contrast to urban and other built environments (e.g. Hartig, Korpela, Evans, & Gärling, 1997; Laumann, Gärling, & Stormark, 2001; Herzog, Maguire, & Nebel, 2003).

Elaborating on the concept of restoration, Kaplan and Kaplan (1989) propose four successive stages that a restorative experience might pass through (pp. 196–197). Starting with “clearing the head” of cognitive clutter, such as distracting thoughts, a person can continue into recovery of directed attention capacity. With the “cognitive quiet” gained in these initial stages of restoration, the person is better prepared to consider “matters on one's mind” that otherwise have gone unheard. Finally, if not disrupted, the restorative experience can entail reflection on one's life, priorities, possibilities, actions, and goals.

These stages are grouped by Herzog, Black, Fountaine, and Knotts (1997) into attentional recovery (stages 1 and 2) and reflection (stages 3 and 4). In a study that used a scenario method to establish theoretically relevant goal-sets, Herzog et al. found that natural environments were preferred over sports/entertainment settings and urban environments by people who had been instructed to imagine needing either to rest directed attention or to reflect on serious personal problems. Natural environments, it appeared, were seen as offering better opportunities for both attentional recovery and reflection. The process that is seen supporting both attentional recovery and reflection is described as soft fascination, which refers to fascination (effortless attention) distributed across aesthetically pleasing features of an environment without a high degree of involuntariness. In contrast, hard fascination involves a high degree of involuntariness, such that the events and objects command attention. In line with the view that hard fascination can support attentional recovery but not reflection, Herzog et al. found that sports/entertainment settings were preferred over urban environments by people with an attentional restoration goal, but not by people with a reflection goal.

Following from ART, a central hypothesis posed in the present study concerns the relationship between attentional fatigue, restoration needs, and preferences for different environments. ART supports the view that the possibility for attentional restoration will guide the environmental preferences of people who are attentionally fatigued. As natural environments may better support attentional restoration than urban environments (e.g. Hartig, Mang, & Evans, 1991; Kuo & Sullivan, 2001), this view allows elaboration on the general finding that natural environments are consistently preferred over urban environments. It implies that the antecedent condition of attentional fatigue should increase preference for natural over urban environments.

Yet an eventual interpretation of results should be based on more than a test of this hypothesis alone, as such a test cannot do more than imply that preference is (or is not) related to restoration. More direct evidence of the restoration motive as a basis for differential preferences is called for. Such evidence would refer to the perceived likelihood of restoration in one or the other environment and the degree to which the person actually favors restoration under the prevailing circumstances. In line with this thinking, we expect that the likelihood of restoration is judged to be greater in natural than in urban environments and that restoration is favored more by attentionally fatigued people than by people who are not attentionally fatigued. Our expectations refer to both recovery from attentional fatigue and reflection.

Tests of these expectations could better validate the notion that preferences for natural environments are grounded in the restoration possibilities they are seen to offer. Validation would be stronger if it was determined that natural environments are less preferred for motives other than restoration, and in particular for motives that are less likely to be fulfilled in natural environments. A comparison of natural and urban environments allows for discrimination in this regard. Whereas people may prefer natural environments for recovery from attentional fatigue and for reflection, they may prefer urban environments for socializing, entertainment, cultural events and resources, and similar motives that as a set we will refer to as social stimulation. This leads us to expect that the social stimulation motive will be more strongly related to preference for urban environments than for natural environments. We expect further that social stimulation is judged more likely in urban than in natural environments, and that social stimulation is favored more by people who are not attentionally fatigued than by attentionally fatigued people. This last expectation follows from the understanding that being in an urban environment requires directing attention to navigation, traffic, avoiding crowds, and so forth which, in urban life, go hand-in-hand with much social stimulation.

Two issues have remained implicit above and will be explicated in this study. First, traditional environmental preference measures do not differentiate with respect to people's behavior in the environments being evaluated. Yet, being in an environment necessarily involves some form of behavior. That people associate particular environments with particular behaviors has been demonstrated in a host of studies. Genereux, Ward, and Russell (1983) showed that 50% of the variance in similarity ratings of places can be explained by the behaviors that are associated with those places. More generally, behavior settings (Barker, 1968; Schoggen, 1989) constitute the interplay of environments and modes of conduct. Thus, preference for an environment should be implicitly, if not explicitly, associated with preference for behavior in that environment. This association is made in various theories about environmental preferences. For example, Kaplan and Kaplan (1989) posit that preference for a scene is partly determined by possibilities for staying oriented and receiving new information while hypothetically roaming around in the inferred third dimension of the picture plane.

Different behaviors may not only influence preference ratings because of the different requirements that these behaviors make on environments. Relevant for this study, different behaviors also require different amounts of directed attention. Contrast a bus driver driving his vehicle through an inner city area during rush hour with a pedestrian going out for a coffee on a nearby city square on a quiet Sunday morning. Because different behaviors make different demands on directed attention, standardizing the behavioral perspective from which a preference rating is given should constitute a methodological improvement. To provide a behavioral perspective, we use preference for walking, in an urban or natural environment, as the dependent variable in this study. Walking is a common type of behavior that is possible in both types of environment, and it ordinarily imposes low demands on physical fitness, abilities, knowledge and directed attention.

A second issue implicit in our previous discussion concerns the way in which specific motives explain variation in traditional preference measures. By representing a particular behavior with our dependent variable, we could address this second issue by using theory on attitudes toward behaviors and the motivational bases of those attitudes. We implemented the expectancy-value model of attitudes, which is a component of the theory of reasoned action (Fishbein & Ajzen, 1975; see also Eagly & Chaiken, 1993). The expectancy-value model states that the attitude toward a behavior is based on (a) the likelihood that a behavior will have specific outcomes and (b) the evaluation of those outcomes. The product of the likelihood of occurrence and evaluation of a specific outcome can be called a motive. The sum of the product of likelihood and evaluation of each of the set of salient outcomes (i.e. the outcomes that people can easily retrieve from memory) constitutes the attitude toward a specific behavior. This is described in the formula: $\text{A}{}_{\mathrm{<mtext>act</mtext>}}\text{=}\text{∑}\text{BE}$, where A_act stands for the attitude toward a behavior, B for belief about the likelihood of the outcome of a behavior, and E for the evaluation of the outcome.

Decomposing an attitude into beliefs and evaluations in this way yields some important advantages over commonly used environmental preference measures. For one, the effects of antecedent condition (e.g. attentional fatigue or its absence) and contextual influences on the attitude can be tracked. Moreover, they can be tracked not only to specific outcomes, but also to key components of outcomes, namely, their likelihood of occurrence and their evaluation. Also, the power of each outcome in determining the attitude can be empirically assessed. As demonstrated by Nisbett and Wilson (1977), people may be poor judges of their own motives. Independent assessment of attitudes and motives allows for investigation of the strength of relationships in an unbiased manner. Taking into account concerns raised about directly interpreting the BE-products (Evans, 1991; Eagly & Chaiken, 1993, pp. 234–236), we use the correlations of beliefs (likelihood of occurrence of an outcome) with attitudes as a measure of the strength of a motive. The evaluation scores are nonetheless useful here for investigating the differential effects of antecedent condition.

In sum, theoretical and methodological insights from research on restorative environments and attitudes toward behaviors can help us to better understand differences in preference for natural and urban environments. Extending earlier work, in this study we examine the role of antecedent condition (attentional fatigue vs. its absence) in shaping preference for an urban and a natural environment. We also examine the strength of relationships of restoration and social stimulation motives with attitudes toward a commonplace behavior in each environment. Furthermore, we test for differences in the judged likelihood and evaluation of restoration (i.e. attentional recovery and reflection) and social stimulation motives for people under different antecedent conditions. Thus, we have specified our expectations about the preference-shaping role of antecedent condition and restoration motives according to the expectancy-value framework of attitudes toward behavior. This leads to the following hypotheses to be tested in this study:

• The preference for (and attitude toward walking in) a natural environment is greater (more favorable) than the preference for (and attitude toward walking in) an urban environment.

• The difference between the preference for (and attitude toward walking in) a natural environment and the preference for (and attitude toward walking in) an urban environment is larger for people who are attentionally fatigued than for people who are not attentionally fatigued.

• The likelihood of restoration, specified as recovery from attentional fatigue and reflection, should be significantly and positively correlated with attitude toward walking in both the urban and the natural environment.

• Both (a) recovery and (b) reflection are judged to be more likely while walking in a natural environment than while walking in an urban environment.

• The evaluation of restoration, that is (a) recovery from attentional fatigue and (b) reflection, is more positive given attentional fatigue than in an absence of attentional fatigue.

• The likelihood of social stimulation is more strongly related to the attitude toward walking in an urban environment than to the attitude toward walking in a natural environment.

• Social stimulation is judged to be more likely while walking in an urban environment than while walking in a natural environment.

• The evaluation of social stimulation by people without attentional fatigue is more positive than by people with attentional fatigue.