UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Affecting cue selection through contextual integration Eichhorn, David P. 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2003-860698.pdf [ 5.77MB ]
JSON: 831-1.0091235.json
JSON-LD: 831-1.0091235-ld.json
RDF/XML (Pretty): 831-1.0091235-rdf.xml
RDF/JSON: 831-1.0091235-rdf.json
Turtle: 831-1.0091235-turtle.txt
N-Triples: 831-1.0091235-rdf-ntriples.txt
Original Record: 831-1.0091235-source.json
Full Text

Full Text

R u n n i n g head: C O N T E X T U A L C U E S E L E C T I O N A F F E C T I N G C U E S E L E C T I O N T H R O U G H C O N T E X T U A L I N T E G R A T I O N b y D A V I D P. E I C H H O R N B . A . , The Univers i ty of Cal i fornia , Rivers ide , 1991 M . A . , The Univers i ty of Br i t i sh C o l u m b i a , 1996 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S Department of Psycho logy W e accept this thesis as conforming to the required s tandard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A A u g u s t 2003 © D a v i d P. Eichhorn , 2003 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT Contextual Cue Selection ii This dissertation presents research examining the role of contextual patterns, salience, and individual differences in the determination of how much is incidentally remembered from a cognitive task performed during the exploration of a naturalistic outdoor environment. Previous empirical findings suggest that the human mind often selects cues for the storage and retrieval of information based upon a rigid, pre-determined hierarchy, frequently disregarding useful contextual cues in favor of features most directly relevant to the information itself. Drawing from environmental psychological principles, factors are outlined that contribute to the salience of contextual cues, the most important of which is the cognitive integration of the context with the observer and the integration of both with the task or mental operation at hand. Such integration is referred to as "contextual integration" and may represent an over-arching schema that serves as a cognitive or affective indicator of personal significance. The first of two reported experiments demonstrated superior memory for contexrually integrated stimuli over those given more rudimentary consideration. The second experiment found changes in memory resulting from an interaction between the type of task performed and the mediating role of a cognitive style known as field-independence. This interaction supports the notion that there are predictable patterns to the cognitive management of contextual information. The effects of these patterns are better accommodated by contextual integration than any single construct such as personal relevance or depth of processing. Furthermore, arousal states and affective ratings of the environment, in Experiments 1 and 2, respectively, showed differential changes in reaction to more or less integrated situations. Conducted almost entirely in a natural environment, the research presented attempts to more closely merge the empirical ideals of the environmental and cognitive areas in psychology. Contextual Cue Selection iii T A B L E OF C O N T E N T S Abstract ii Table of Contents iii List of Tables v List of Figures vi Acknowledgements vii Introduction 1 Perspectives on the Selection of task-relevant and Environmental Cues . . . . 4 Cognitive Approaches to Integration 6 Order Out of Chaos: The Rational, Adaptive M i n d 6 Associative Network Approaches 8 Memory in Context 10 A Broader Reality: The Environmental Perspective 14 Contextual Integration and the Present Empirical Approach 18 The Factors 20 The Relations 21 The Present Empirical Approach 25 Cognitive Aspect 26 Affective Aspect 27 Trait Aspect 28 Experiment 1 29 Method 33 Subjects 33 Materials 33 Equipment 36 Design and Procedure 37 Results 46 Cognitive Measures 46 Verbal Memory 46 Spatial Memory 47 Affective Measures 48 Mood and Environment Ratings 48 Discussion 52 Cognitive Implications 52 Affective Implications 53 Experiment 2 54 Method 56 Subjects 56 Materials and Equipment 56 Design and Procedure 56 Results 57 Contextual Cue Selection iv Cognitive Measures 57 Verbal Memory 57 Spatial Memory 58 Affective Measures 58 Mood and Environment Ratings 58 Discussion 61 Cognitive Implications 61 Affective Implications 63 General Discussion 65 References 71 Appendix The Word Lists and their Semantic Ratings 77 Contextual Cue Selection v LIST OF T A B L E S Table 1. Mean ratings of mood in Experiment 1, by levels of task and time 49 Table 2. Mean ratings of mood in Experiment 2, by levels of task and time 60 Contex tua l C u e Selection v i LIST OF FIGURES Figure 1. I l lus t ra t ion of potent ia l first-order interactions 19 Figure 2. The relat ions be tween the person (P), his or her env i ronment (E), and the task at h a n d (T) 20 Figure 3. The M o o d G r i d 35 Figure 4. The des ign for Exper imen t 1 40 Figure 5. Images f rom the computer s imula t ion 41 Figure 6. Reca l l b y levels of task and context reinstatement, Exper imen t 1 47 Figure 7. E n v i r o n m e n t a l pleasantness b y levels of t ime and task, Exper imen t 1 . . . 51 Figure 8. The des ign for Exper imen t 2 55 Figure 9. Reca l l b y levels of task and C A B - C f group, Expe r imen t 2 59 Figure 10. Spat ia l m e m o r y b y levels of task, Exper iment 2 59 Contextual Cue Selection vii A C K N O W L E D G E M E N T S This research was supported by an American Psychological Association Dissertation Research Award, without which the more technical aspects of the project would have been impossible. A n immeasurably large amount of gratitude is due to Peter Suedfeld for his unwavering patience and support throughout this project. Many thanks also to Lawrence Ward, Don Wilkie, Eric Eich, Stanley Coren, Mark Schaller, Dan Perlman, Robert Gifford, and Colin Campbell for their invaluable comments and suggestions on earlier drafts of this dissertation. Thanks also to the Brandlmayrs for getting me outdoors and helping me feel at home no matter my environment. Contextual Cue Selection 1 I N T R O D U C T I O N Both individual and species survival are in large part determined by the ability to successfully, and selectively, process the highly complex and changeable stimulation comprising the typical environmental context. Evolutionary and sociobiological theorists talk about environmental pressures (Wilson, 1975), many perceptual and environmental psychologists inspired by the work of Gibson (1979) refer to "affordances," but it does not take any special knowledge or perspective to understand that there is an inescapable influence of environments upon everything that occurs within them. Despite its ubiquity and irrefutable role in human activity and development, the influence of the environment upon human cognition and behavior is difficult to isolate and quantify. That is why most psychology research is conducted in controlled laboratory settings and customarily does not consider the role of environmental or other contextual factors. There are notable exceptions, including but not limited to the entire field of environmental psychology and some perspectives in cognitive psychology. In fact, researchers from nearly every field of psychology make at least occasional attempts to incorporate their findings into a broader applicable context, and greater generalizability is an implicit goal of all psychology research. The present analysis focuses upon research approaches that make regular use of explicit manipulations and measurements of contextual factors and their interrelations in attempts to determine those qualities that render an environment, or other contextual element, more or less salient, meaningful, or personally relevant. The principal idea behind the present analysis is that one of the most important qualities of real environments is their potential to satisfy the human need to perceive relations among the elements of any situation. That is, humans have an innate need to integrate their surroundings into a meaningful internal representation (Kaplan & Kaplan, 1983) and it should be possible to predict reactions resulting from the success or failure of that integration process. For instance, new or unusual elements introduced Contextual Cue Selection 2 into a familiar environment will attract increased attention to that environment (Nadel & Willner, 1980; Jacobs, Thomas, Laurance, & Nadel, 1998) as the new element is integrated into the schemata representing the entire contextual milieu. Depending upon the perceived importance of integrating the new element, the inability to do so can lead to increased arousal and eventual frustration or even stress (Stokols, 1979), or the diversion of attention away from the anomalous information if it is deemed unimportant or unmanageable. This is the first sense of the term "contextual integration" that will be used throughout this thesis, the degree to which elements of a situation work together to form an interrelated conceptual schema. Used as a verb, the term can also refer to the human tendency to cognitively restructure information in an effort to create such a state of affairs. The third, and most frequent, use of the term contextual integration in this analysis will be in reference to the application of that ideal to research investigating human-environment relations. The general concept of integration is certainly not new and has re-emerged in psychology with increasing frequency over the last several decades, particularly in the areas of cognitive and environmental psychology. Addressing a conference on levels of processing in memory, Jenkins (1979) acknowledged an already rapidly growing trend by saying "Everyone now knows that memory phenomena are much more complicated and contextually determined than we used to think they were" (p. 430). In his arguments, it is clear that Jenkins uses the term "contextual" in a general sense to mean that a broad range of factors bear on memory and that as many as possible should be considered at a time. Physical context is just one subset of these possible sources of variation. Context-dependent memory research focuses almost exclusively on physical contexts (Godden & Baddeley, 1975; Smith, 1988; Smith & Vela, 2001), as well as internal contexts such as mood (Blaney, 1986; Bower, 1981; Eich & Metcalfe, 1989), in what is possibly the best example of a consistent integrative approach to research in cognitive psychology. Contextual Cue Selection 3 The relation between a person and his or her environment is merely the starting point for research in environmental psychology, which takes an holistic approach to psychology wherein complex experimental integration involves multiple interacting elements. For example, environmental psychologists interested in the cognitive effects of a stressor in a particular environment (e.g., noise in a school setting, Ward & Suedfeld, 1973) will make every effort to analyze a situation very much like the one in question (preferably in the actual setting, as in the Ward and Suedfeld study) rather than simply relying upon standards for the stressor obtained in controlled laboratory situations. This situational realism is also applied to the tasks performed during an experiment. Environmental psychologists try to have subjects perform tasks that are normal for an environment, using variations in performance on those tasks as broader indicators of the effects of environmental factors, rather than using only physiological or simple yes-no response indicators of a reaction. As would be expected, the greater resulting generalizability is often accomplished at the expense of such experimental hallmarks as internal validity and replicability (Stokols, 1987), but the reactions of subjects in such studies are less likely to involve demand characteristics or response biases than those of subjects in laboratory settings. Reactions can also be more pronounced, as in the case of a frustrated subject in the Ward and Suedfeld (1973) study who used a fire axe to chop the cable supplying recorded highway noise outside his classroom. This dissertation describes theory and research that support a cognitive approach to environmental psychology. A principal concern of such an approach is the t o maintenance of experimental rigor and replicability in real or realistic environments without losing the qualities of those environments that determine important factors such as their affective significance or personal meaning. Existing research will be reviewed that has already defined some of the more important determinants of such environmental qualities, and issues related to their measurement. Relevant findings and assumptions in cognitive psychology are presented to establish the concept of task-Contextual Cue Selection 4 relevant cue selection and its relevance to the process of mentally structuring the environment, the importance of which is reviewed through the literature on environmental interaction and preference. To help bring these issues together, contextual integration is presented as a means of characterizing situations in terms of the degree to which they represent an optimal combination of elements that might satisfy the human need for an ordered context. Research is presented that uses changes in memory performance and mood to examine the role of contextual integration in the day-to-day process of organizing and acting upon environmental information. From an evolutionary perspective it is reasonable to assume that not all stimuli are equal, that some stimuli that are of functional importance to the organism will have special affective properties associated with them (Kaplan, 1987, pp. 4-5). Perspectives on the Selection of task-Relevant and Environmental Cues If certain patterns of contextual information can acquire more meaning and significance than others, with related changes in attention and other cognitive functions, then a better understanding of the patterns might provide a template for more narrowly focused analyses of the functions both antecedent and consequent to such affective responses. As the preceding quote from Kaplan (1987) suggests, there may be patterns of stimulation to which the human perceptual system is particularly sensitive, to which it allocates greater cognitive resources. The quote does not mention patterns explicitly, but as an environmental psychologist, Kaplan typically writes about complex stimuli such as landscapes and no doubt would agree with the present argument that patterned stimuli are most relevant to discussions of cognitive and affective processes. It is unlikely that any single stimulus would develop and later re-establish anything like a significant affective response without a corresponding interpretive context. A rock on its own may simply be a rock, but if holding it, feeling it, and seeing familiar streaks through its color draw forth from memory an earlier experience wherein the rock was found and became the center of a stimulating conversation on a significant day, then Contextual Cue Selection 5 that rock is elevated in meaning to the status of a personal keepsake, a treasure. Its simple features call forth the entire pattern of stimulation of which it was a part when it was first encountered, a remembered pattern that lends the rock special meaning and significance. This is by no means a new idea. The entire souvenir trade in tourist locations throughout the world capitalizes on this type of contextual meaning and place association. Jacoby and Craik (1979) discuss the same relativity and context-dependence of an item's significance in their discussion of distinctiveness and its contribution to the depth of stimulus encoding. According to them, "meaning is a set of contrasts resulting from distinctions required when interpreting the item in the context of some task" (Jacoby & Craik, 1979, p. 3). What is of particular interest in the present analysis is the effect of this meaning ascription upon later processes such as attention and selection, in addition to the aforementioned concern with the underlying organizational nature of the ascription process itself. In a description of the role of complex contextual patterns in emotional reactions and their consequent effects upon cognitive activity, Russell (2003) includes "all the information possessed about the external cause" (p. 148) in a long list of factors contributing to "Core affect [which] is part of the information used to estimate affective quality and thus is implicated in incidental acquisition of preferences and attitudes" (p. 149). He also mentions mood-congruent memory, the idea that what is remembered from a stimulus array may depend upon the correspondence of its emotional valence with the mood of the individual. Similarly, in what is widely accepted as the best explanation of a long, complicated history of context-dependent memory results, Eich (1995) provides evidence that affective states associated with reinstated learning environments mediate the beneficial effects upon memory for material learned in those environments. Given these strong, complex associations among personal significance, cognition, and the environment, the contextual integration approach provides a framework for research that considers all three. That framework allows a more Contextual Cue Selection 6 narrowly focused delineation of the patterns of contextual information that elicit significant affective responses of the type likely to influence cognitive processes. Eich's (1995) manipulation of mood, independent from changes in the environment, can be seen as a contextually integrated approach. Not only did it provide a broader, more multi-faceted perspective on the factors contributing to context-dependent memory effects, and did so using a natural environment, but at a more basic level it also demonstrated the empirical ability to systematically vary a naturally occurring person-environment relation. In the area of environmental psychology, the research conducted by Ward, Snodgrass, Chew, and Russell (1988) also embodies what can be interpreted as a contextually integrative perspective. Their experiment manipulated complex relations among contextual elements using meaningful tasks that entailed distinct "plans," and measured their effects upon both environmental affective ratings and memory tests. The distinctiveness of the different task plans created meaningful associative relations among the person, the task, and the experimental situation, which is interpreted under the present scheme as a fairly high degree of contextual integration. Further analysis of these examples, their relation to personal relevance and its role in cue selection, and the possible mechanisms involved will follow more logically after a review of other existing integrative approaches to the organization of human thought and its relation to patterns of contextual information. Cognitive Approaches to Integration Order Out of Chaos: The Rational, Adaptive Mind. As mentioned earlier, Jenkins (1974,1979) makes some important arguments for a "contextual" approach to research in psychology, which should entail the careful consideration of a comprehensive spectrum of variables influencing an experimental task. In particular, he proposes a "tetrahedron" of interrelated variables that converge upon ongoing cognitive activity. These are the "orienting task," the "materials," the "criterial tasks," and the "subject" variables (Jenkins, 1981, p. 233). At first, these appear to differ only slightly from the Contex tua l C u e Selection 7 key factors p roposed b y the present analysis, those of the task, the environment , and the person. Sp l i t t ing the task variable into subject- and task-related sub-components is a fair ly m i n o r theoretical divergence, but by l i m i t i n g his "contextual" focus to the task-related materials and to subject variables such as expertise, Jenkins (1981) makes on ly pass ing reference to the t remendous p o o l of relevant in fo rmat ion offered b y the phys ica l env i ronment i n w h i c h the task is performed. A l t h o u g h his m o d e l is not contextual i n the sense most impor tant to the present analysis, Jenkins (1981) does emphasize the impor tance of cons ider ing the interactions a m o n g the categories of variables, par t icu lar ly i n his earlier explanat ion of the m o d e l (1979), w h e r e i n he expl ic i t ly defined each interaction. A n d e r s o n (1990; A n d e r s o n & Schooler, 1991) pays par t icular attention to the importance of env i ronmenta l cues, suggest ing that the selectivity of the h u m a n cognit ive sys tem m a y be the result of adapt ive pressures th roughout the evo lu t ion of the species. C o g n i t i v e systems are thereby at tuned to the envi ronments i n w h i c h they develop so that resources are not was ted o n attempts to incorporate u n l i k e l y events. The basic idea is that at any poin t i n time, memories va ry i n h o w l i ke ly they are to be needed and the m e m o r y system tries to make avai lable those memor ies that are most l i k e l y to be useful. The m e m o r y system can use the past h is tory of use of a m e m o r y to estimate whether the m e m o r y is l i k e l y to be needed n o w (Ande r son & Schooler, 1991, p . 400). In other w o r d s , wha t the m i n d elaborates u p o n and eventual ly stores i n l o n g term m e m o r y is not o n l y wha t it th inks it w i l l need, but that w h i c h it can easily integrate into exis t ing knowledge . Thus, A n d e r s o n introduces one par t icular ingredient w h i c h makes a l l the difference i n attempts to unders tand h u m a n dec is ion and cue selection strategies: the rat ional , adapt ive nature of the h u m a n m i n d , w h i c h is u n i q u e l y capable of seeking out that w h i c h is famil iar and understandable i n an otherwise chaotic d i sp lay of env i ronmenta l and task-related s t imul i . Schacter (1999) concludes a recent r ev iew of theories of m e m o r y and its l imita t ions b y suppor t ing the A n d e r s o n and Schooler (1991) perspective, a n d b y w a y of i l lus t ra t ion prompts the reader to consider "what w o u l d be the consequences and costs of re ta ining the m y r i a d of contextual details that define our Contextual Cue Selection 8 numerous daily experiences?" (Schacter, 1999, p. 197). Clearly there is much to be gained in terms of processing efficiency if cognitive processes can be simplified as much as possible, and much to be gained in terms of predictive power if psychology can delineate systematic ways in which the mind accomplishes that simplification. A n understanding of the hierarchy behind the cue selection process is an important step toward that delineation. Associative Network Approaches. One of the most highly organizational approaches to cognitive psychology is the associative network theory of memory, for which the article by Collins and Loftus (1975) is probably the most often cited. The Collins and Loftus model describes a process by which a stimulus activates a number of "nodes" in a semantic knowledge network, which in turn activate related nodes and so on in parallel until enough of the proper connections are made to constitute a representational match to the original triggering stimulus. Once a matching pattern of internal representation is obtained, the target stimulus is recognized as having been encountered before. The aspects of the network approach that are most relevant to the present analysis are: (a) that relatively simple conceptual constructs that exist in the mind can associatively combine to form larger, more elaborate ideas, and (b) that an external stimulus can give rise to some sort of mental activation which in turn can trigger an infinite variety of thoughts and associations by traveling different paths through this network of differentially related concepts. The applications for, variations upon, and extensions of the network theory are numerous and, due to their intuitive appeal, propagate swiftly (see Forgas, 1999, for a recent review and extension). In the approach most relevant to the present analysis, Bower (1981) discusses the way in which a mood or emotion can activate an entire network of related memories, concepts, and sensations, as well as the inhibitory effect an unrelated mood can have upon the attempt at activation. "The different emotion node will call up interfering associations that will compete with recall of the correct Contex tua l C u e Selection 9 target items" (Bower, 1981, p . 136). B y consider ing compet ing associations, B o w e r acknowledges the potent ia l for associative connections to impede as read i ly as assist the act ivat ion of a target representation. H e also used the ne twork m o d e l to describe the influence of emot ion u p o n several cogni t ive processes i n add i t i on to memory, inc lud ing selective attention and social perception. In a l l of these instances, conceptual schemata i n l ong te rm m e m o r y are considered the basis for de te rmin ing the relat ive impor tance of i n c o m i n g s t imula t ion i n order to pr ior i t ize the a l locat ion of attention to those s t imul i . "The contents of consciousness are the sensations, concepts, and propos i t ions whose current act ivat ion leve l exceeds some threshold" (Bower, 1981, p . 134). Thus ne twork theory suppl ies an explanatory basis for the present contextual in tegrat ion approach w h e r e i n the key contextual factors of the person, the task, and the env i ronment m a y represent p r i m a r y nodes responsible for o rgan iz ing basic associative patterns f rom w h i c h a l l further p ropos i t iona l analyses m a y proceed. The hierarchical organiza t ion and h i g h degree of interact ion detai led b y the models def in ing them not on ly make in tu i t ive sense, they also serve w e l l as explanatory constructs for a fair amount of data. In an early descr ip t ion of h o w associative theories account for cue effectiveness, A n d e r s o n and B o w e r (1973, pp . 341-343) report the results of a s tudy done b y a student of theirs (Wanner, 1968). In the experiment, s ingle w o r d s were better sentence retr ieval cues w h e n the target sentence was one of a matched pai r conta ining more anci l lary associations to the cue w o r d than if the target was the sentence w i t h fewer associations. This is on ly one of m a n y such exper imental results A n d e r s o n and B o w e r (1973) exp la in w i t h an associative ne twork approach. Similar ly , the w o r k s of C o l l i n s and Loftus (1975) a n d B o w e r (1981) are f i l led w i t h such examples, most of w h i c h fal l unde r the general category of research into "pr iming ," w h i c h is a greater tendency for an i t em to be either recal led or otherwise p roduced i f it, or mater ia l related to it, has been active i n consciousness recently. W i t h examples such as these, and m u c h more empi r i ca l suppor t across a w i d e var ie ty of subtopics i n p sycho logy i n the years since, the associative ne twork theory of m e m o r y Contex tua l C u e Selection 10 has p r o v e n its usefulness as a power fu l predict ive tool. T ryon (2002) suggests that the ne twork approach is capable of integrat ing theories f rom numerous areas i n psycho logy and related fields. The far-reaching appl icabi l i ty of ne twork theories attests to the general usefulness of cons ider ing complex interactions a m o n g variables w h e n s tudy ing cogni t ive and related phenomena. M e m o r y i n Context . The s tudy of m e m o r y i n context has been the most active cogni t ive psycho log ica l approach to the invest igat ion of h o w env i ronmenta l and task variables interact to affect an ind iv idua l ' s thought processes a n d performance. For this reason, a n d to more closely match Eich 's (1995) study, the present experiments were des igned to a l l o w a context-dependent approach to the m e m o r y measures. The te rm "contextual m e m o r y effects" w i l l be used throughout this dissertation to refer to the general f ami ly of m e m o r y improvements , decrements, a n d congruencies attributable to the change or constancy of external and / or in ternal contextual cues present at the time a m e m o r y is encoded and again w h e n it is retr ieved. The proto-typica l example for this class of phenomena is the phys ica l env i ronment man ipu la t i on used b y G o d d e n a n d Badde ley (1975), w h e r e i n m e m o r y performance was lowest w h e n the env i ronment (in this case underwater or o n land) d u r i n g retr ieval d i d not match the s tudy envi ronment a n d highest w h e n the environments matched. Th i s specific effect is referred to as place-dependent memory . S imi la r effects have been found for pharmacological-state contexts (Eich, Weingartner, St i l lman, & G i l l i n , 1975; E i c h & B i r n b a u m , 1982) as w e l l as mood-state contexts (Bower, 1981; E ich , 1995; E i c h & Metcalfe, 1989). These effects are referred to col lect ively as state-dependent memory. W h e n referr ing general ly to place- and state-dependent memory , it is customary to use the te rm context-dependent memory . The even broader t e rm "contextual memory" is used here because this also can inc lude mood-congruent memory, w h i c h was described earlier as the selective m e m o r y for mater ia l cor responding to one's current m o o d . Contex tua l C u e Selection 11 A l m o s t f rom its earliest incept ion, contextual m e m o r y research has been more interested i n the memory-re la ted dec is ion processes i n v o l v e d than i n m e m o r y performance per se. A t first, contextual m e m o r y researchers were p reoccup ied w i t h the variety and k i n d of contextual cues available to the typ ica l exper imental subject. Smi th , Glenberg , and Bjork (1978, p . 352) men t ion t ime of day, color of wal l s , and appearance of the experimenter as just a few examples of the m a n y potent ia l ly impor tan t contextual features that a subject might use as m e m o r y cues. Th i s perspective, w h i c h w i l l be referred to as the cue-combinat ion, or exact-sum, approach, considers any failure to replicate contextual m e m o r y effects a reflection of the inab i l i ty to capture the exact combina t ion of contextual features necessary to generate a useful or salient contextual cue. It q u i c k l y became apparent, however, that the mere presence or s imple combina t ion of m e m o r y cues does not guarantee their con t r ibu t ion to m e m o r y performance. A mul t ip le-example failure to replicate the phys ica l context-dependent m e m o r y effect was reported b y Fernandez and Glenberg (1985) and s imi la r contradictory results have appeared i n the state-dependent l i terature (Bower & Mayer , 1989), a l l ind ica t ing the necessity for more complex explanations. A c k n o w l e d g i n g the role of person-related factors, S m i t h (1979, p . 464) refers to the "cuing" and "strategy" hypotheses to compare the general abi l i ty and the decision, respectively, of an exper imental subject to u t i l i ze phys ica l contextual cues to retrieve an associated m e m o r y or interrelated group of memories . B y cons ider ing the dec is ion processes of the i n d i v i d u a l , this emphasis u p o n cue selection contributes another piece to the overa l l p u z z l e of contextual memory, and comes closer to the present integrative ideal , but s t i l l ignores the importance of classifying the types of person-task interactions invo lved . E i c h and his colleagues (Eich, 1980; E i c h et al., 1975; E i c h & Bi rnbaum, 1982) were a m o n g the first to use an i m p l i c i t l y hierarchical system to exp la in h o w inconsistent contextual m e m o r y effects migh t result f rom the differential relevance attributable to the var ious cues associated w i t h a g iven task si tuation. F r o m their perspective, failures to demonstrate the influence of context u p o n m e m o r y m a y reflect the presence and Contex tua l C u e Selection 12 selection of more salient cues, such as related w o r d s or category exemplars (among the items referred to b y E ich , 1980, p . 160, as "list" and "copy" cues). This emphasis u p o n the compet i t ion a m o n g task-relevant cues has been dubbed the "outshin ing hypothesis" (Smith, 1988, p . 19) and has been used most often to exp la in the consistent resistance of categorized list members and recogni t ion m e m o r y to the influence of contextual cues. C o m p a r e d to the contextual ly integrative idea l advocated i n the present analysis, the exact-sum approach places too m u c h emphasis u p o n i n d i v i d u a l features of the context, of w h i c h there m a y be dozens. A n experiment c o u l d precisely match every imaginable phys i ca l feature across times of s tudy and test, but this w o u l d amount to a m a n i p u l a t i o n of on ly one key exper imenta l factor: the envi ronment . The cue selection and ou t sh in ing hypotheses fare better i n that they classify potent ia l cues according to their salience, i m p l y i n g that factors such as meaningfulness and personal relevance m a y contribute to the cue selection process, an essentially integrative approach. That is to say that, b y suggest ing the role of o rganized self-referent dec is ion factors i n the process of cue selection, these approaches i m p l i c i t l y acknowledge the impor tance of consider ing complex person-environment relations i n analyses of the ongo ing contro l of cognit ive activity. Person-envi ronment relations signify the interact ion be tween o n l y two of three key factors that interact to determine the three impor tant relations that a fu l ly integrated approach mus t consider. The th i rd key factor is the task to be performed, and it interacts w i t h the other factors to create the other t w o p r i m a r y relations, between the task a n d the.person and between the task and the envi ronment . A n analysis is not fu l ly contextually integrated unless it considers the complex interrelations a m o n g the three key factors that interact to determine the ove ra l l context of any cognit ive activity. For instance, an experiment that independent ly manipula tes on ly one or two of these factors or relations at a t ime is l i ke ly to p roduce inconsistent results. F r o m this broader perspective, the abi l i ty of a famil iar category to "outshine" an envi ronmenta l cue i n a place-dependent m e m o r y experiment can be exp la ined as a Contex tua l C u e Selection 13 person-task re la t ion (the association of list members w i t h famil iar categories) out-sh in ing a single env i ronmenta l cue, as opposed to the less informat ive explanat ion of a task-related cue d r a w i n g more attention than an environment-re la ted cue. A s an example of the importance of the interrelations a m o n g the key contextual factors, consider that i n the G o d d e n and Baddeley (1975) context-dependent m e m o r y experiment a l l the participants were experienced S C U B A divers , a fact necessitated b y the use of an underwater environment . This impl ies a close re la t ion be tween the person and the env i ronment that has not been replicated i n most subsequent research. A more precise def in i t ion of "close relat ion" w i l l emerge later i n this argument, but i n the most basic sense, wha t is meant is the fami l ia r i ty or personal relevance that comes f rom repeated interact ion w i t h an envi ronment (as w i t h S C U B A d i v i n g or other place-specific behaviors) or t h rough learned or constructed place attachment (as w i t h re l ig ious homelands or i m a g i n e d locations). This personal relevance is the same type of affective response discussed earlier i n relat ion to Russell 's (2003) analysis of "core affect." The existence of such a re la t ion w i l l increase or decrease the subject's use of the env i ronment as a contextual cue, depend ing u p o n the nature of the other relations. A close re la t ion be tween any number of person and env i ronment factors m a y increase the usefulness of reinstat ing the s tudy context at the t ime of test, but a co-occurr ing strong re la t ion be tween the person and the task (e.g., a familiar, demand ing , or in ternal ly focused task) c o u l d counteract, overr ide, or even magni fy the resul t ing contextual m e m o r y effects. This is an impor tant concept for the present analysis, and not w i thou t precedent, as i l lus t ra ted b y the earlier K a p l a n (1987) quote and others to be presented and discussed later, i nc lud ing the fo l lowing: Informat ion that aids i n m a k i n g sense out of the env i ronment is l i ke ly to be par t icu lar ly salient. Information that a l lows an i n d i v i d u a l to make more accurate inferences about his whereabouts s h o u l d be h i g h l y v a l u e d (Kaplan , 1975, p . 93). Wi thou t s t ray ing f rom the contextual m e m o r y research, w e see the impor tance of these relations descr ibed b y Fernandez and Glenberg (1985, p . 344): Contextual Cue Selection 14 We interpret and remember specific events as being embedded within a structure of more global (and more local) events interrelated by physical causes and psychological goals of the actors taking part in the events. Also, the environmental context is usually closely related to the events that occur within the environment. That is, the environment may be perceived as causing the event or enabling the event to occur. Thus, it is likely that relations between the environmental context and events are integral to the representation of naturally occurring events. The key to understanding the significance of these relations and their interactions lies in treating the three basic contextual components as categories of activation, the interaction of which culminates into a contextually integrated, and thereby more memorable, experiential whole. The resulting affective response may, in fact, be the very type Eich (1995) posits as mediating context-dependent memory effects. A Broader Reality: The Environmental Perspective Environmental psychology has successfully adapted cognitive psychological principles to suit the broader reality it seeks to define. While the fully "contextual" ideal heralded by Jenkins (1979,1981) remains the exception in cognitive psychology, such integrative approaches are not only central to environmental psychology, they are oftensupercededby those embracing "transactionalism," which adds the variable of passing time to the integrative approach (cf. Altman & Rogoff, 1987, for a thorough review of various world views in psychology). Most relevant to the present analysis, the environmental area takes a bi-directional perspective on person-environment relations that provides particularly fertile theoretical ground for further exposition of the contextually integrated approach to research in psychology. Though he was technically a perceptual psychologist, the work of Berlyne (1960, 1963) established one of the first groups of explanatory variables destined to form the foundation of thought and practice in environmental psychology. He delineated four variables, novelty, uncertainty, conflict, and complexity, which he referred to as "collative variables" (Berlyne, 1960). These variables define classes of stimulus properties that exert varying attentional demands upon the human perceptual system. In 1963 Berlyne demonstrated the utility of a subset of these variables in the prediction of Contex tua l C u e Selection 15 s t imulus selection. Subjects presented w i t h more and less i r regular ly patterned s t imul i spent more t ime exp lo r ing the more i r regular (MI) s t i m u l i than the less i r regular (LI) s t imul i , but were more l i ke ly to rate L I s t imu l i as "pleasing". Ber lyne exp la ined these f indings i n terms of the processing demands of the s t i m u l i and the arousal leve l experienced at var ious points d u r i n g a typ ica l s t imulus exposure pe r iod . Typically, ini t ial presentation is associated w i t h an increase i n arousal as the perceiver anticipates and then receives the p rev ious ly u n k n o w n material . F r o m that momen t of i n i t i a l exposure there is a g radua l decrease i n arousal, and any concomitant discomfort or unease, as the perceiver becomes more famil iar w i t h the part icular s t imulus array. Supposedly, d u r i n g this t ime the m i n d of the perceiver is ass imila t ing as m u c h of the p r o v i d e d image as possible, b reak ing it into subsets of itself that are more l i k e l y to match exis t ing conceptual representations i n the l o n g te rm m e m o r y ne twork . Therefore, the more famil iar or easily processed elements there are compr i s ing the s t imulus array, the less t ime the perceiver needs to process it and therefore the faster the recovery f rom the in i t i a l increase i n arousal . "Characteristics such as order ly spatial arrangement, coherent g r o u p i n g of elements, repeti t ion, and redundancy, w h i c h d i s t inguish our L I patterns, appear to curb the in i t i a l rise i n arousal and to accelerate recovery f rom it." (Berlyne, 1963, p . 288) The increasing tendency to rate the L I patterns as "pleasing" is thus attributed to the perceivers ' satisfaction at h a v i n g met the processing demands q u i c k l y and relief at the swift reduc t ion of arousal afforded b y the patterns. B y def in ing and measur ing the influence of s t imulus properties u p o n h u m a n in format ion process ing a n d preference, Ber lyne (1960, 1963) set the precedent for future analyses of the cogni t ive significance of env i ronmenta l features. F r o m the env i ronmenta l p sycho logy perspective, the role of cogni t ion i n person-environment interact ion has been examined most extensively b y Stephen a n d Rache l K a p l a n . F o l l o w i n g a s imi la r attempt b y W o h l w i l l (1968), K a p l a n , K a p l a n , and Wendt (1972) conducted an experiment i n w h i c h subjects p r o v i d e d preference and complex i ty ratings for slides of env i ronmenta l scenes. To i m p r o v e u p o n the W o h l w i l l (1968) study, K a p l a n Contextual Cue Selection 16 et al. (1972) made an effort to create and pre-divide their slides such that they depicted distinctively "natural" contrasted with "urban" visual scenes. Their results indicated that complexity predicted preference for the scenes within, but not across, the natural and urban categories. Subjects greatly preferred the natural scenes, but apparently not for any reason directly related to complexity, despite the role of complexity in preference ratings within the categories. The hierarchical nature of decision making processes, discussed earlier in the context of cognitive psychology, begins to show itself here. Although complexity is a powerful predictor of environmental preference, its inability to account for the distinction between environments independently rated as natural and urban suggests that there are more salient features exerting greater influence upon the decision processes responsible for preference. , In a review of several approaches to environmental preference, Kaplan (1987) proposes four predictive variables: complexity, mystery, coherence, and legibility of a given visual scene. He classifies these variables in terms of two dimensions, the first being.their contribution to one or the other of "two affectively important informational outcomes", which he defines as "'Understanding' (comprehending or making sense of a scene) and 'Exploration' (being held by the setting, being attracted by or pulled toward sources of additional information)" (Kaplan, 1987, pp. 10-11). The other classification dimension Kaplan uses is the difference between information that is "immediately available" and that which is "predicted or promised" (Kaplan, 1987, p. 11). In more recent work (Kaplan, Kaplan, & Ryan, 1998), this latter dimension has been further clarified as the difference between "2D," immediately perceptible features, and "3D," features that would require movement through the scene to be fully appreciated. Particularly noteworthy is Kaplan's (1987) explanation of the predictive variables in terms of adaptive pressures and the evolutionary advantage of a natural inclination to extract as much information as possible from the environment. These themes, and a cognitive psychological perspective, also guide the book by Kaplan and Kaplan (1983) in which Contex tua l C u e Selection 17 they p r o v i d e a thorough analysis of the h u m a n endeavor to better unders tand and funct ion i n the phys i ca l environment . The evolu t ionary and information-processing perspectives in t roduced b y the Kap lans have con t inued to influence envi ronmenta l preference research and theory. H e r z o g a n d h is colleagues have p roduced numerous recent articles (e.g., H e r z o g & F lynn-Smi th , 2001; H e r z o g & K u t z l i , 2002; H e r z o g & Mi l l e r , 1998; H e r z o g & Shier, 2000) cover ing a large array of env i ronmenta l scene types and perspectives. Throughou t these studies, factors such as complex i ty and mystery continue to demonstrate s trong effects u p o n judgments and classifications of the scenes. There have been some refinements, such as the incorpora t ion of factors f rom comparable theories (e.g., Apple ton ' s , 1984, "refuge" factor) and the def ini t ion of add i t iona l predictors, such as "Vis ib i l i ty" and "Locomotor Access" (Herzog & K u t z l i , 2002), but a l l w i t h i n the same general theoretical pu rv iew. Peron, Purce l l , and their colleagues consider the w o r k of the Kap lans , but also inc lude factors such as representational schemata and changes i n frames of reference to clarify otherwise inexpl icable changes i n preference w i t h i n and between scene types (Peron, Purce l l , Staats, Falchero, & L a m b , 1998; Purce l l , L a m b , M a i n a r d i Peron, & Falchero, 1994; Purce l l , Peron, & Berto, 2001). The i r use of frames of reference bears some resemblance to the use of plans i n the W a r d et al . (1988) study, w i t h s imi la r effects u p o n the outcomes, and therefore represents another example of a fair ly integrative approach to contextual relations. Fur ther d i scuss ion of the relevance of the env i ronmenta l perspective on psychologica l processes mus t awai t the more detai led descr ip t ion of contextual integrat ion i n the next sub-section. In general, as part of its empi r i ca l s tandard of m u l t i -leve l analyses, env i ronmenta l psycho logy offers a f ramework for p red ic t ing cogni t ive reactions to par t icular classes of envi ronmenta l s t imula t ion that migh t frui t ful ly be app l i ed to the analysis of cue selection strategies. U s i n g the k n o w n extent of h u m a n processing capabili t ies as a guide, A n d e r s o n (1990; A n d e r s o n & Schooler, 1991), the Kap lans (Kaplan , 1975,1987; K a p l a n & K a p l a n , 1983; K a p l a n , K a p l a n , & R y a n , 1998), Contextual Cue Selection 18 and environmental psychologists since have demonstrated that it is possible to carefully examine the pattern of a stimulus array and assess whether its component features will influence approach, avoidance, selection, or any of a host of other behavioral and cognitive outcomes. The most valuable aspect of environmental psychology is not its attention to the role of environmental features in human activity, it is the field's careful attention to the entire milieu of person-environment interdependencies that makes it a promising field of scientific inquiry. Emphasizing the importance of empirical attention to relations more than elements, Suedfeld (1991) reminds us that we must examine "experiences within environments" rather than just people or environments in isolation. The transactional approach of environmental psychology has made for some attractive theories, but has proven difficult to implement. Attempting to account for and control so many variables and all their potential interactions is typically not practical. That is why it is so important to have a means of determining what components of an overall experience contribute most to its overall meaning for an individual. With its classification of contextual relations into three primary categories, the proper activation of which defines an elaborate personal event, a fully integrated lasting schema in an associative network, the contextual integration approach may significantly increase the practicality of transactional analyses. Contextual Integration and the Present Empirical Approach It was emphasized in the section on contextual memory that the concept of contextual integration involves more than the arbitrary combination of memory cues hypothesized in most network models. Such models suggest that, regardless what features are combined, as long as they are related in some way, there is some specific number or valence of cue combinations that will trigger memory retrieval. Figure 1 illustrates the complex interactive potential that emerges from just a small subset of person and environment features and the relations that may exist among them. For instance, the receptivity of the person to naturalistic stimuli, arising from a confluence Contextual Cue Selection 19 Person Features Environment Features Personality Complexity Mood Coherence Motivation Familiarity Arousal level Naturalness Experience Importance Figure 1. Illustration of potential first-order interactions. of the person features listed in Figure 1, could interact with the more naturalistic features of an environment to render them more memorable or personally significant than any less naturalistic stimuli in the setting. This figure bears a strong resemblance to Metcalfe Eich's "illustration of the convolution of two items", a conceptualization of the way in which two items converge into "a higher order associative trace." (Metcalfe Eich, 1982, p. 631) The similarity between the figures highlights the similarity between the views, which focus on the network or pattern of activity stimulated by isolated combinations of cues and the power of past experience to enhance or impede that activation. Contextual integration places less importance on specific cues and paths of activation, focusing instead upon the categories of activation to which cues belong and the division of those categories into the three key contextual factors of the person, his or her environment, and the current task, activity, or mental routine being performed. The central premise of the present analysis is that it is not just the activation of cues belonging to each of these categorical factors, but a measure of interaction among them, that is necessary to elevate an overall momentary experience to the status of a unique, distinctive event. That is, full contextual integration may be a necessary Contextual Cue Selection 20 precursor to the attribution of affective properties to a given experiential milieu. Such full contextual integration represents the highest level in what is treated as a multi-level model of the process by which stimulus information is matched and integrated with existing knowledge. Figure 2 represents the interactive potential among the three key contextual factors. It is important to remember that the same complexity of interaction between the two key factors illustrated in Figure 1 (relation "a" in Figure 2) also exists between the person and the task and between the environment and the task (relations "b" and "c" in Figure 2). Although the empirical application of this model may at first appear overwhelmingly complex, it actually reduces complexity by grouping the interactive components into generalizable categories of activation. The remainder of this sub-section will elaborate the predictive potential of this approach. E Figure 2. The relations between the person (P.), his or her environment (E), and the task at hand (T). The Factors There is really no need to describe the key contextual factors at great length. Each is capable of infinite variety, but there is nothing mysterious about them and they are relatively self-explanatory. The only possible exception is the task factor which is C o n t e x t u a l C u e Selection 21 interpreted rather b r o a d l y so as to i n c l u d e not just observable activities, b u t m e n t a l routines as w e l l . There are t w o i m p o r t a n t points that need to be m a d e about the role of the i n d i v i d u a l factors i n the o v e r a l l m o d e l . First , it is i m p o r t a n t to note that despite the complex i ty w i t h w h i c h a single contextual factor m a y be dealt, whether i t is considered across t ime, v a r i e d i n appearance, or r e m o v e d entirely, it is s t i l l just a single factor. F o r instance, the s tandard e n v i r o n m e n t a l m a t c h / m i s m a t c h procedure u s e d i n place-dependent m e m o r y studies represents the m a n i p u l a t i o n of the s ingle factor of the e n v i r o n m e n t . State-dependent m e m o r y studies are m a n i p u l a t i n g the p e r s o n rather t h a n the e n v i r o n m e n t , but this is s t i l l just one independent factor, unless the e n v i r o n m e n t is cons idered part of the m o o d m a n i p u l a t i o n . Second, there is n o real place i n the current m o d e l for sol i tary modif icat ions of s ingle contextual factors a n d the arbitrary c o m b i n a t i o n of contextual features w i t h o u t regard for the p r i m a r y factors to w h i c h they b e l o n g can y i e l d unrel iable results. The p r o p o s e d k e y contextual factors m a y p r o v i d e a structure for more precisely categoriz ing contextual influences, a n d yet their organizat iona l strength is considered subs id iary to, a n d dependent u p o n , a considerat ion of their interrelations. The one possible exception to the ru le against the e x a m i n a t i o n of single contextual factors is the p e r s o n factor. G i v e n that the p e r s o n is the factor of greatest interest a n d complexity, it m a y be sensible to a l l o w the considerat ion of i n d i v i d u a l components of the p e r s o n regardless of whether or not they are part of a larger interrelated set of contextual factors. Sex a n d personality, for instance, are t w o person factors that can y i e l d a great deal of i n f o r m a t i o n relative to the ease w i t h w h i c h they can be i n c o r p o r a t e d into a p s y c h o l o g y experiment. The Relations There are certain m i n i m u m relations a m o n g the k e y contextual factors that exist regardless the exper imenta l m a n i p u l a t i o n or everyday s i tuat ion. The p e r s o n is a lways i n some e n v i r o n m e n t , the p e r s o n is a l w a y s p e r f o r m i n g some task (or at least cogni-Contextual Cue Selection 22 tively processing, in a non-task-oriented situation), and the task is always present against the same environmental backdrop as the person. The integrative relations in the present model, however, are meant to indicate something more substantive than these basic co-occurrences. To use an analogy, simply tossing flour, butter, and milk into a pan and adding heat will not produce a smooth bechamel sauce. The correct proportion of ingredients must be combined at the proper time and with the appropriate amount of heat. In a similar fashion, full contextual integration can only be achieved if there is a specific pattern of interaction among contextual factors which exceeds the minimal level implied by mere co-existence. As for the "heat" that binds the contextual "ingredients," the affective response to a contextually integrated situation is the most likely candidate. Returning to the cognitive psychology literature, person-environment relations ("a" in Figure 2) have received the most empirical attention, primarily from contextual memory research. Changing the environment from the time of study to the time of test does not really involve a "relation" between the person and the environment, but the multiple study room manipulation used by Smith (1979,1984) came closer by drawing attention to the change. As mentioned earlier, the pre-existing person-environment relation in the Godden and Baddeley experiment (1975) also approaches more closely the level of integration implied by the present approach, but the relation received no special attention in that study. It was not until Eich's (1995) study that the relation between the person and the environment was given proper consideration in contextual memory research. Eich's (1995) experiments raise the level of contextual integration by independently manipulating mood and the presence of a pleasant natural environment. The result was a more comprehensive explanation for context-dependent memory effects than any that has been offered before or since. There are also empirical examples of the other primary relations in the contextual memory literature. Eich and Metcalfe (1989) provided an example of the independent manipulation of the relation between a person and his or her task (represented by line "b" in Figure 2) when they had subjects either generate or read the to-be-remembered Contextual Cue Selection 23 stimuli. Eich (1985) manipulated the relation between the task and the environment (relation "c" in Figure 2) when he instructed subjects to either mentally integrate the stimuli with environmental features or to simply form independent images of them. Both experiments found significant differences in recall depending on whether the study and test contexts were matched or mismatched, but only for the integrated stimuli in the case of the latter study. To explain the lack of context-dependent effects for non-integrated stimuli, Eich (1985) suggests that the independent imagery instructions may have diverted subjects' attention away from the environment, which might otherwise have contributed to what he suggests was an underlying "experiential context dependence" (p. 769) for material integrated with the environment. This argument anticipates Eich's (1995) later use of the more contextually integrative approach of independently manipulating mood and environment. Working at the level of "higher-order, composite concepts (e.g., person-environment fit, place identity)" (Stokols, 1987, p. 51), environmental psychology "should attempt to discover the 'meanings' of the events to participants" (Altman & Rogoff, 1987, p. 34). After the person-environment relation, the most common relation considered in environmental psychology is that between the person and the task. A classic example is the independent manipulation of predictability and subject control over aversive stimuli in the determination of the stressful effects of noise (Glass & Singer, 1972). Similarly, Baum and Valins (1979) determined that degree of control, and learned helplessness resulting from little or no control, mediate the perception of crowding in a university dormitory environment. Summarizing the environmental perspective on general stress research and theory, Evans and Cohen (1987) name a variety of such factors that might mediate a person's reaction to an environment. It is arguable that these factors can be interpreted as variations in the single contextually key factor of the person, rather than as representatives of a more broadly meaningful contextual relation. The former position would render these mediating factors no more theoretically significant than individual differences. They are Contextual Cue Selection 24 considered examples of the important person-task relation from the contextual integration perspective, however, because they significantly alter the person-environment relation such that the meaning of the environment is changed for the person as a result. In sum, "the meaning of a physical configuration of an environment has powerful influences on whether those physical conditions will elicit stress" (Evans & Cohen, 1987, p. 596; emphasis added) or whatever reaction is under consideration. At the very least, these mediating factors should be seen as person factors that exert a significant influence upon, or derive their meaning from, the person-environment relation. Whether comprising one of the key contextual relations or simply modifying another, the inclusion of mediating factors such as control, evaluation, or receptivity in models of human reactions to, and functions in, the environment still only raises those models to a moderate level of overall contextual integration. Stokols (1979) also considers such models incomplete. He suggests greater complexity is the key to better representing true environmental contexts, tempered by a consideration of the plans and goals a person brings to a situation. In a later discussion, Stokols (1987) says that experimenters must more precisely determine the "effective context of the target phenomenon" (Stokols, 1987, p. 46). The point about plans and goals (Stokols, 1979), regardless whether in relation to more or less complexly representative environments, is exactly in line with the contextual integration approach. According to this approach, it is in fact the plans given to subjects that render the situation in the Ward et al. (1988) study more representative of a true context. This is because the plans guiding behavior had demonstrable effects upon the meaningfulness of the task-environment relation. Theoretically, if the over-arching experience to which an individual is exposed obtains the highest level of contextual integration, all other influences upon memory or any other cognitive process should become secondary. In the case of memory, recalling any one part of a contextually integrated whole should call forth the full experience, which is why contextual memory is of such interest when addressing this topic. Context Contextual Cue Selection 25 reinstatement should not matter in the case of full contextual integration because the stored information is maximally available regardless the testing situation. High level, or full, contextual integration is, like transactionalism, a rather high empirical ideal that is difficult to achieve in practice and therefore quite rare. In some instances, as with stress research (Evans & Cohen, 1987), obtaining the same impact as the situation of interest may in fact be unethical. As Stokols (1987) says, environmental psychologists must do what they can to match contexts as much as possible, focusing upon those features that matter most. Contextual integration may offer a useful template for categorizing those features and determining the pattern of contextual information most likely to trigger a genuine affective reaction. Once aware of the full contextually integrated situation, the mind establishes an array of reactions appropriate to potential changes in the given milieu. According to Ward and Russell (1981b), when a situation reaches this level of significance, recognizability, or meaning, it is said to have developed "emergent properties" that, in part, help to categorize the situation as part of an existing conceptual framework for environmental meaning. That framework guides what can be expected from, and contributed to, an environment in a given category. What this type of environmental categorization suggests with regard to the dependence of affective states upon cognitive processes is arguable, but in terms of the receptivity and reactivity of the human mind to environmental stimulation, it clearly suggests that there are many levels of environmental influence and that the mind has a hierarchical priority structure for dealing with them. The use of a fully integrative contextual approach may be the best possible way to track and predict the function of that hierarchy and truly begin the process of learning what matters to the human mind. The Present Empirical Approach Given the foregoing argument that a primary function of human cognition is to create order out of life's continual chaos, it is possible to see the naturally occurring Contextual Cue Selection 26 contextual integration process as a sort of mental or affective triangulation. In order to make sense of an environment, and to aid the tasks that must be accomplished in it, the mind seeks meaningful associations and prefers, or at least begins with, those that fall into the relational categories, or key factors, described earlier. In a sense, full contextual integration represents the lowest common denominator or base template for the complex network of associations comprising an individual's personal knowledge. Incoming stimuli that do not fit the contextually integrated pattern are either rejected as anomalous but irrelevant or tagged for further analysis to resolve the idiosyncrasies. In order to capture this mental triangulation process and to allow some predictions regarding what types of contextual cues are likely to be used in which situations, the present empirical approach simultaneously addresses three different aspects of the experiments in this project. The cognitive aspect is concerned with the creation and assessment of the desired contextual relations, the affective aspect examines subjects' reactions to the cognitive elements, and the trait aspect considers some person-related factors that might mediate the effects of the cognitive and affective elements. Cognitive Aspect. The foremost concerns in the present experiments were the creation of realistic and modifiable relations among the different contextual categories and the use of unobtrusive outcome measures to gauge the salience of those relations for the subjects. Subtle variations in the instructions were used to create task manipulations like those in the Ward et al. (1988) study, engendering the same types of meaningful associations between the tasks and the environment. In a similar way, the element of stimulus choice was used to alter the relation between the subjects and their task, in essence changing the depth of their consideration of the materials. Finally, a context-dependent-memory approach was used in the design of the experiments in order to capitalize on previous demonstrations of the person-environment relation and its possible sensitivity to cue selection strategies. The dependent measures of verbal and global spatial memory were the primary means by which the present experiments assessed what was most salient about the Contextual Cue Selection 27 experimental situations. Memory is often used as an experimental tool to reflect the function of deeper mental processes which cannot be measured directly themselves. By tracking what is successfully retained in memory, we are observing the preservation of that which is important to the cognitive system that achieved the retention. It is in this way that memory for a particular stimulus pattern may be understood as an indication of an underlying association to that pattern, an association that may represent something akin to personal relevance. In general, it was hypothesized that an increase in the number of activated contextual relations would increase the salience of, and memory for, contextual elements both verbal and spatial, but only if each activated relation involved a different pair of contextual categories. Multiple activations of a single categorical relation were not expected to contribute to greater memory for the elements involved. Redundant person-environment relations, however, were recognized as a possible exception in that spatial memory should benefit from the resulting increased attention to environmental elements. With regard to context-dependent memory effects, the performance of subjects experiencing the minimum and maximum levels of contextual integration was expected to reflect little or no influence from a reinstated context. The context was not expected to be salient enough in the minimal condition and the high overall performance in the maximum condition was expected to leave little room for improvement. Context-dependent memory effects were considered most likely to occur in situations with incomplete contextual integration but involving at least one task-environment relation, due to the resulting increase in the relevance of the environment to the task. Affective Aspect. In addition to the purely cognitive measures, the present experiments included affective measures to further assess potential changes in the salience of contextual elements attributable to their incorporation in activated relations. There were not any direct manipulations of mood, as in Eich's (1995) experiment, but it Contextual Cue Selection 28 was carefully tracked, and affective and other qualitative ratings of the environment were also collected. The general hypothesis for the affective ratings was an overall increase in positive ratings and a decrease in negative ratings associated with increases in contextual integration. This follows from the assumption that higher levels of integration should render a context more cognitively manageable and therefore more appealing, as measured by environmental and personal affective ratings. As Kaplan (1975) suggests, "humans would be expected to prefer an environment where both recognition and prediction can be achieved without undue effort" (p. 98). Trait Aspect. The design of the present experiments did not allow for the independent manipulation of person-related factors, but it was considered worthwhile to control for two factors deemed relevant to the other topics, sex and cognitive style. Although neither is a "trait" in the often-used sense of a personality factor, this term was chosen as a category label because of its general reference to stable, as opposed to transient, individual qualities. There were no sex differences expected with the verbal memory measures, nor with the affective measures, but sex was balanced across groups in the present experiments because past research suggests there may be sex differences in spatial abilities (McGuiness & Sparks, 1983; Moffat, Hampson, & Hatzipantelis, 1998). Based upon the results of these studies, male subjects were expected to perform slightly better on the spatial memory measure. The cognitive style measure used in the present experiments is also related to spatial ability. A n attempt was made to balance the groups with regard to subjects' scores on a measure of field-independence (Witkin, Oltman, Raskin, & Karp, 1971), which is roughly defined as the tendency to readily separate an object from its background or context. Sternberg and Grigorenko (2001) debate whether it is a tendency or an ability, perhaps even a component of intelligence, but there is general Contextual Cue Selection 29 consensus that field-independence is not a personality factor due to its basis in observable perceptual performance differences. Assessing field-independence was considered important for the present study because there is research to suggest that field-independent individuals may be more inclined to engage in "cognitive restructuring" of incoming stimulation (Davis & Frank, 1979). Cognitive restructuring can take the form of three different possible operations, all capable of influencing performance in the present experiments: (1) breaking up the organization of a stimulus complex so that its elements can be operated upon separately or in new combinations; (2) providing structure for an ambiguous stimulus complex; and, (3) providing a structure different from that implied by the inherent structure of the stimulus complex (Davis & Frank, 1979, pp. 469-470). The evidence for field-independence-related differences in memory is somewhat inconsistent and has been summarized by Goodenough (1976) as the result of differences in the use of memory strategies rather than in general ability. Smith (1984) found differences in his contextual memory results attributable to field-independent versus field-dependent characteristics, although he did not discuss the result much because it did not interact in any meaningful way with his other results. Given the variability in the designs and results of previous research, there was no attempt made to predict any role of field-independence in the present experiments, hence the decision that it would be sufficient to balance the groups with regard to this measure. EXPERIMENT 1 Ideally, the contextually integrated experiment should represent a more significant, interesting, and enjoyable event for those participating in it. Such attributions undoubtedly would contribute to more ecologically valid results given that such affective indicators are probably some of the criteria that people use to determine their willingness to participate in day-to-day activities. Both of the present experiments were designed to demonstrate the cognitive and affective effects of complex contextual Contextual Cue Selection 30 relations in a situation more natural but less easily controlled than that of the typical laboratory, using methods intended to be as engaging yet unobtrusive as possible. Thus, every effort was made to create experimental situations that were as realistic and enjoyable as possible. The choice of a naturalistic outdoor environment in which to conduct the experiments was just the first step of that process. The Nitobe Japanese Garden on the University of British Columbia campus was chosen because (a) it is a naturalistic environment, which makes it more engaging and pleasant, (b) it is only 6 minutes away, by foot, from the University of British Columbia Psychology Department, and (c) it was the pleasant mood environment in Eich's (1995) experiment. The qualitative guideline for a realistic, engaging experimental situation was also applied to the creation of the primary task for both experiments, which was designed to make sense in the environment, to complement its design and appearance. To use Barker's (1968,1987) terminology, the task was intended to feel appropriate for the behavior setting. The primary task created for the present research was essentially a wayfinding task wherein subjects were told to explore the Garden as thoroughly as possible. This task was chosen because it is not an unusual behavior in a new outdoor environment, particularly one with inviting paths winding about it, like the Nitobe Garden. An exploration task was also chosen because it would easily accommodate the additional tasks related to the cognitive and affective aspects of the experiments. The secondary tasks, intended to manipulate the various levels of contextual integration, control for alternative processes, and present the to-be-remembered stimuli, required considerably more compromise between the qualitative guideline of realism and the quantitative reality imposed by the desired level of experimental rigor. To satisfy the intended quantitative guideline for experimental precision and control, and to fit within a general incidental verbal memory paradigm, all the tasks needed to be only slight modifications of the same basic task and they had to present word lists to the subjects without letting them know that they would be tested for memory of the words later in the experiment. The similarity across tasks was to ensure that any elicited Contextual Cue Selection 31 differences in performance would be attributable to the different levels of contextual integration each task invoked, as opposed to some other quality, such as difficulty or required depth of processing (Cermak & Craik, 1979; Lockhart & Craik, 1990). This requirement often conflicted with the qualitative guideline determining that the tasks seem as natural and unobtrusive as possible. The incidental nature of the memory task fortunately satisfied both the qualitative and quantitative guidelines. Qualitatively, incidental memory is a far closer approximation to the function of memory in day-to-day circumstances, since most people do not go about life trying to memorize everything they encounter. Quantitatively, incidental memory techniques are preferable to intentional memory techniques because the latter afford subjects the opportunity to utilize their own rehearsal strategies, greatly reducing experimenter control over the actual mental processes elicited by the tasks (Bellezza & Reddy, 1978; Lockhart & Craik, 1990; Neill, Beck, Bottalico, & Molloy, 1990). Therefore, the basic secondary task in both of the present experiments was creative storytelling. Although this task and the changes used to vary its integrative nature were somewhat unusual and not always probable for the Garden location, they were deemed satisfactory because (a) in all but the most contextually integrated conditions, discordance between the task and environment was appropriate, (b) task-environment discordance supported the ostensible role of the secondary task as a distraction intended to interrupt the subjects from their primary task of exploring the environment, and (c) all tasks met the minimum guideline for contextual integration, they each appeared to acquire their own unique meaning and significance. There is ample research demonstrating that convincing instructions can instill what has been called a cognitive or perceptual set that can alter affective judgments (Ward & Russell, 1981a), memory for the environment (Ward et al., 1988), environmental categorization (Genereux et al., 1983), and visual perception (Coren, Porac, & Theodor, 1986). Care was taken in the present experiments to ensure that it was exactly this type of distinct approach that was adopted for each task. Contextual Cue Selection 32 A n ancillary purpose of these experiments was to demonstrate the valuable role of computer control and automation in research examining such complex phenomena. Al l procedures in both experiments, from group selection through presentation of the instructions to data collection, were initiated and controlled automatically by computer programs written specifically for this project. In the simplest terms, these experiments both involved the presentation of word lists interspersed with the exploration and rating of an outdoor environment, along with mood ratings, all occurring on the first of two days. Testing for memory of the word lists and environmental locations occurred 24 hours later, give or take an hour, with or without some form of reinstatement of the first day's environmental context. Both experiments were run relatively concurrently with the exception that Experiment 1 was nearly half completed before Experiment 2 started and was finished about a month earlier. The experiments were run continually over the course of a year and a half. The weather varied according to the seasons throughout this time, but extreme conditions were always avoided such that it was never raining and there was only snow on the ground twice. Other variations were minimal and distributed across all groups. The principal goal for the first experiment was to test the predictive power of the contextual integration approach, compared with the less specific cue-combination approaches more typical in psychology, such as in context-dependent memory research. If a meaningful, memorable association with the features of a new situation can result from any random activation of cues, regardless of their quality, nature, or source, then activating three redundant versions of the same relation in the contextual integration model should exert no less influence upon cognitive and affective measures than the same number of activations distributed across the three different relations of the model. Contextual Cue Selection 33 Method Subjects The subjects in the first experiment were 54 men and women, ranging in age from 18 to 28 years old, randomly chosen from among those University of British Columbia psychology undergraduates responding to a departmental advertisement. Al l subjects received extra course credit for their participation. Sex and field-independence were balanced across all groups. The most common paper-and-pencil measure of field-independence, the Embedded Figures Test developed by Witkin et al. (1971), was deemed too time-consuming to administer, so a related measure, the five minute "flexibility of closure" (Cf) subscale of the Comprehensive Abilities Battery (CAB; Hakstian & Cattell, 1975), was used to infer this cognitive style. The C A B - C f was administered after group assignment, however, and exhibited a pronounced ceiling effect, so the balance across groups was imperfect. A l l subjects were screened when they signed up in order to exclude those who had visited the Garden before, so as to eliminate any pre-existing familiarity. Three subjects were allowed who had been there before, but so long ago and for such a brief time that they insisted they had no recollection of the Garden's features. Materials The to-be-remembered stimuli were 60 common nouns, chosen from the Toglia and Battig norms (1978; clusters 7 and 8). The word list was created using another computer program designed for this experiment, but applicable to any research wishing to use verbal stimuli from the published norms from which the program draws. The program can be instructed to generate lists of any length, to restrict the length of the words, and to generate lists weighted on any of the seven "semantic dimensions" reported in the norms (Toglia & Battig, 1978). The only restrictions placed on the words for the present list were a minimum word length of four letters, a maximum word length of eight letters, and an average concreteness rating of 6.3 (out Contextual Cue Selection 34 of a possible 7). After slight manual adjustments were made in order to limit internal associations (e.g., reducing the unusually high number of food-related words), the words in the final list had an average concreteness rating of 6.0, an average imageability rating of 5.7, and an average meaningfulness rating of 4.4. This list was then divided randomly into six different lists of ten words each. The lists, complete with semantic dimension ratings, can be found in the Appendix. A l l subjects were shown all ten words from each list but were instructed to use only half of them for the verbal processing (e.g., storytelling) task. In conditions where a person-task relation was desired, the subjects were allowed to choose their own words. Subjects in the other conditions received the words chosen by another subject, in a yoked control fashion, and were told that they were random selections. The logic behind having some subjects choose their own words was that the additional processing required to select the words would create a more meaningful, and therefore more memorable, personal association with the chosen words, in a way similar to the effect of generating words from ambiguous stimuli (Slamecka & Graf, 1978). The yoking process of giving the word choices from each subject in a word-choice group to a single subject in each of the other groups in the same experiment was intended to off-set the fact that each chosen list was unique. Mood was assessed at the beginning and end of each experimental day using the Positive and negative affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) and the mood grid developed by Russell, Weiss, and Mendelsohn (1989). The P A N A S is comprised of 20 adjectives that subjects rate on a five point scale to describe how well each one indicates their mood, from "very slightly or not at all" to "extremely". The mood grid, shown in Figure 3, was developed from a circumplex approach to the measurement of affect (Russell & Lanius, 1984) and is completed by placing a single point on the grid to indicate mood. That point is then converted into two scores, ranging from -4 to 4, indicating arousal level and the qualitative assessment of that arousal level in terms of how pleasant or unpleasant it feels. Contextual Cue Selection 35 E X T R E M E L Y H I G H A R O U S A L m x H 3J m 2 m r -< "D r m > > z H E X T R E M E L Y L O W A R O U S A L Figure 3. The Mood Grid. Subjects were also asked to use the P A N A S and the mood grid to rate the experimental environments each day. This was not peculiar in the case of the mood grid, which was adapted from a method for interpreting environmental ratings (Russell & Lanius, 1984), but there is no known precedent for using the P A N A S in this way. The most similar approach is that of Ward et al. (1988) who used a 40-item questionnaire designed more specifically for rating environments. For both measures used in the current study, subjects were told to rate "the mood attributes of the environment itself," with emphasis placed on the independence of those attributes from their own moods. For clarification, they were told to rate "how this environment would generally affect the average person." Subjects also rated the environments for complexity and preference, on a 5-point scale ranging from "little or no" to "very." Prior to these ratings, subjects were given the Herzog, Kaplan, and Kaplan (1976, p. 630) definitions for these variables: "intricateness, or the opposite of simplicity" for complexity, and "how much you like the scene for whatever reason" for preference. In addition, it was emphasized that complexity should be considered a visual or aesthetic measure of the environment Contextual Cue Selection 36 as a whole, as opposed to an approximation of something more scientific like the biological complexity of individual plants. Equipment A l l of the computer automation and presentation was accomplished using Apple's HyperCard program. HyperCard uses the HyperText computer language, an early precursor of HyperText Markup Language (HTML), which is currently the most widely accepted language for creating web-sites. HyperCard's use of HyperText, however, is far more powerful than H T M L , allowing variable naming, complex mathematical calculations, and conditional responses to input. The HyperCard programs designed for these experiments were presented on an Apple Power Macintosh G3 computer with a 17-inch monitor. A hand-held Canon ES970 8mm portable video recorder was used to record images and short videos of the Garden for later use in the spatial location memory test. The video camera was also used by subjects as part of the exploration of the environment on the first day of both experiments. The videos for use in the location memory test and a short demonstration simulation of the Nitobe Garden were digitized using the Adobe Premiere video digitizing and editing program. They were compressed into Apple QuickTime files using the Sorenson 3 compressor. Still image editing was accomplished with Adobe Photoshop and a Wacom Graphire pen tablet. A l l word presentation and data collection occurring in the Nitobe Garden used a Palm Hie personal data assistant (PDA) ruruung the HanDBase data entry and management program. Because the data on the P D A could be transmitted directly to the computer, this effectively eliminated experimenter data entry error, automated the data tabulation process, and greatly reduced the amount of paper used. Better still, it increased the ease and efficiency of data collection and allowed the word selection yoking manipulation described earlier. Without the PDA, the yoking procedure would have been unwieldy and error-prone at best. In addition to the word lists, there were Contextual Cue Selection 37 PDA versions of the PANAS, the mood grid, and all the environmental ratings. Finally, the PDA was used to make shorthand notations of subjects' progress through the garden and it recorded the start and finish times of that progress. Design and Procedure The satisfaction of the cognitive aspect of the present empirical approach required the manipulation of two principal independent variables, the type of processing applied to the stimulus words on the first day (task) and the presence or absence of the first day's environment on the second day (reinstatement). The task variable manipulated the degree of contextual integration by combining various types of processing according to the categories of contextual information activated by them. The reinstatement variable was included to track potential context-dependent memory effects. The three levels of the task variable were independent activities, redundant activities, and integrated activities. The first and third conditions represent the lowest and highest possible levels of contextual integration in this project. In the independent activities condition, subjects were instructed to create a brief story using the five list words pre-selected from the original ten. This, like all the verbal tasks, was repeated six times using each of the six lists, interspersed throughout the exploration of the environment on the first day of the experiment. The independent activities task was not intended to activate any contextual relations, but the person-environment relation was considered to be activated across all conditions by the environmental exploration task. Thus the independent activities condition featured an activation of only the person-environment relation. This was presented to the subjects as an exploratory task that would be randomly interrupted by a storytelling task. At the opposite extreme, the integrated activities task condition was designed to do more than just activate each of the relations depicted in Figure 2. Conceptually, it combined two of the relations (those of the person and environment with the task, Contextual Cue Selection 38 labeled "b" and "c" in Figure 2) and then had the third relation (that of the person with the environment, labeled "a" in Figure 2) interact with that combination. Subjects were instructed to select and then imagine hiding five of the ten words in spots of their choosing in the immediate vicinity. Again, this was repeated six times throughout the Garden. The person-task relation was activated by allowing subjects to choose their words, the task-environment relation was activated by the process of hiding the words, and both were encompassed by the person-environment relation activated by the Garden exploration task. The interruptive nature of the verbal task in relation to the exploratory task was de-emphasized for subjects in this condition in favor of a more interactive interpretation of the relation between the two. For instance, subjects were told to think of the hide-spot selection process as an extension of the exploration of the Garden, a means of more thoroughly examining it. It was assumed that the person-environment relation would achieve greater activation in this condition than in the others, engendered by the subjects' freedom to choose the hiding spots and the spatial orientation and planning required to perform the entire operation, along with the filming of the entire process. This may not have been the case, however, as the spontaneous reactions of most subjects, in all conditions, suggested that at least the verbalization part of the word-hiding task was very much a distraction from the exploration and enjoyment of the environment. The redundant activities task, involving three different activations of the person-environment relation but none of the others, was designed essentially to increase the salience of the environment without increasing its relevance to the verbal task conducted within it. The idea was to bring together the important qualitative and procedural aspects of the distinctly different independent and integrated activities tasks, combining the complexity of the latter with the low contextual integration of the former. In terms of its presentation to the subjects as two competing tasks, exploration interrupted by storytelling, the redundant activities task was nearly identical to the independent activities task. Subjects used pre-selected words to create a story which Contextual Cue Selection 39 they then told aloud. Prior to the verbal part of the task, however, subjects performing the redundant activities task were instructed to do something very similar to portions of the integrated activities task. They were told to imagine they had to hide something in the Garden and to videotape three distinct views of the immediate surroundings in which that could be done. This provided the second activation of the person-environment relation in addition to that provided by the mere exploration of the environment. The third activation of the relation was accomplished at the very beginning of the exploration of the Garden. Upon first entering the Garden, subjects were directed to the bench just inside the gate and the ratings of the environment were done prior to exploration, rather than afterwards as in all other conditions. This immediate consideration of its qualities was expected to increase subjects' overall attention to the environment. The levels of the reinstatement variable were none and physical. In the "none" condition, subjects did not return to the garden on the second day of the experiment and no attempts were made to remind them of it. Half of the subjects were assigned to this reinstatement condition, the other half experienced the physical reinstatement condition, which meant returning to the garden on the second day. The experimental design is summarized in Figure 4, which illustrates the intersection of the three task conditions with the two reinstatement conditions and the number of subjects successfully run through each resulting group. What follows is a description of the actual sequence of events for the experiment. The laboratory room where each subject began the experiment was a small (approximately 4 meters square) non-descript room, the only distinguishing feature of which was that it did not have any windows. The door joining the room to the rest of the laboratory, via access to a common corridor, was left open as often as the noise level in the rest of the laboratory permitted. This was nearly all of the time. The room was illuminated with soft lighting and a floor fan was running at all times to prevent the room from becoming too stuffy and to mask distant noises. Contextual Cue Selection 40 Level of Contextual Integration (Task) Independent Redundant Integrated Activities Activities Activities n — 9 n = 9 n = 9 n = 9 n — 9 n = 9 z o n n 3 f t 3 n 3 Figure 4. The design for Experiment 1. The first day began with the administration of the C A B - C f to determine subjects' cognitive style, during which the experimenter left the room to limit distractions. This was followed by a thorough instruction period, guided by the computer but elaborated by the experimenter, that took approximately half an hour. This included descriptions and practice/baseline versions of the mood measures, in addition to the detailed instructions for the cognitive tasks to be performed in the Garden. Not only was care taken to ensure that the instructions were clear and that they defined a unique approach to the tasks involved, they were followed by a short computer simulation which allowed subjects to practice their given task before proceeding to the actual environment. Using only the area just outside the Garden gate, the simulation allowed subjects to explore several different views and allowed the interaction appropriate for the condition to which they had been assigned. The most important feature of this is that for the subjects in the integrated activities condition, the simulation actually allowed them to physically hide the words in spots in the images that opened up, via animated image pieces, to allow the insertion of the words. Subjects were not just told to imagine hiding words in the Garden, they were given a visual example of the process. Critical stages of the simulation are illustrated in Figure 5. Note that the simulation utilized the labels "word 1" through "word 10" rather than actual words in order to avoid proactive interference from this practice task. Contextual Cue Selection 41 Garden Demo ; 13 S Figure 5a. Ar r iva l at an interactive location in the experimental simulation. Garden Demo I B Figure 5b. A selection disappearing into a hiding spot within the interactive mode. Contextual Cue Selection 42 Upon arrival at the Garden, subjects were given details about the exploration of the Garden (e.g., travel restricted to the counter-clockwise direction) and shown how to use the video camera. The time was recorded on the videotape and the P D A and exploration commenced. For the redundant activities subjects only, the exploration was interrupted almost immediately in order to collect the environmental ratings at the point of initial exposure to the Garden. The six locations at which the experimental tasks took place were all pre-determined to ensure that the words were processed in the same specific surroundings across all groups. Arrival at the locations was orally announced to the subjects in all conditions except the independent activities condition. To avoid an implicit association between the words and the locations (activation of the task-environment relation), arrival at the locations was announced to the independent activities subjects via an alarm on the P D A . They were led to believe that the alarm was sounding at random intervals, when in actuality the experimenter was activating it surreptitiously as each pre-determined location approached. For the redundant and integrated activities subjects, the first step upon reaching the task location was to choose and videotape three unique views of the location wherein they could hide something. The "something" was not specified for the redundant activities subjects, but the integrated activities subjects were told that these were the views wherein they would be hiding their words. The second step for these two conditions, and the first for all other conditions, was to read all ten of the list words aloud. This procedure was partly included to insure that all subjects were familiar with the words (and allowed the explanation of unrecognized words), but was more important as a means of getting all subjects to process all ten words, something required of subjects having to choose from among them but easily avoided by subjects given a pre-selected subset of five words. After the words were either chosen or "randomly" determined, subjects performed the task appropriate to their condition. Subjects in the independent and Contextual Cue Selection 43 redundant activities conditions were given as much time as they liked to create a story and were then instructed to recite the story aloud with the video camera recording. Subjects never visually appeared in the videos but their voices were recorded to keep track of the types of stories told and to motivate subjects to take the task seriously. Subjects in the integrated activities condition also had the video camera running while performing their mental hiding task, with the added instruction to aim the video camera at each of their hide spots in turn while giving a verbal description of the reasoning behind their hiding decisions or some similar verbalization relating each word to its hiding spot. This was intended to match the verbalization of the subjects telling the stories. It also served to underscore the associative process. Upon completion of their respective experimental tasks, all subjects resumed their exploration of the Garden until the next experimental location was reached. Upon arrival back at the front gate, having visited all six experimental locations, subjects in the independent and integrated activities conditions completed the environmental and mood ratings on the PDA. The redundant activities subjects only completed the mood ratings at this point because they had done the environmental ratings earlier. Regardless when it was done, the environment was rated first using the PAN AS and then the mood grid. The personal mood ratings used the same measures in the same order. For the independent and integrated activities subjects, the environment was then rated for complexity and preference in that order, the time for the next day was confirmed, and the first day was complete. As with the other environmental measures, the preference and complexity ratings of the redundant activities subjects were collected earlier. The point at which the next day's time was confirmed was also the first time either the subject or the experimenter learned whether the next day would take place in the Nitobe Garden or back in the laboratory. Both subject and experimenter were kept blind to the subject's exact condition by having the computer place that information in a file on the PDA, which was not consulted until the end of the first day. Contextual Cue Selection 44 The procedures for the second experimental day varied considerably according to the different types of study context reinstatement involved. Subjects in the physical reinstatement condition returned to the Nitobe Garden and performed all tasks on the P D A while sitting on or standing near a bench just inside the front gate, the same location from which the affective ratings had been made the previous day. This bench afforded an expansive view of most of the central Garden, rendering it a perfect spot from which to survey the Garden while reminiscing about the previous day's activities there. Most of the general areas where the verbal tasks were performed could be seen in the distance, but detailed features were not discernible for any but one of the specific locations. Subjects in the "none" reinstatement condition reported to the laboratory and completed the experiment on the computer. Subjects completed the same environmental and mood ratings conducted at the end of the first day. The order of the ratings was counter-balanced such that half of the subjects rated the environment first, whereas the other half rated their moods first. This was to control for any influence of one set of ratings on the other. It is important to ,. note that subjects in the "none" reinstatement condition, returning to the laboratory on the second day, were told to make their environmental ratings with regard to the immediate laboratory surroundings. After the mood and environment ratings, all subjects were asked to venture a guess regarding the purpose of the experiment. This was included to determine if any subjects had anticipated the verbal memory test, which would ruin its incidental nature. Free recall was measured to assess the degree to which the task manipulations affected memory for the task materials. Subjects were told to write down as many of the selected words from the previous day as they could recall, with no penalty for guesses. They were told to continue until they felt they would be unable to recall any more. A time limit was not imposed because it was felt that time pressure would have highly variable effects upon performance, depending upon subjects' reactions to being tested. The instructions implied, however, that five minutes was a reasonable amount of time Contextual Cue Selection 45 to spend on the task, none of the subjects spent longer than 12 minutes and, though not measured, the average time spent was approximately five minutes. When subjects first indicated that they were finished, the experimenter gently encouraged reconsideration by saying "Are you sure? There aren't any more words just on the tip of your tongue dying to get out?" This elicited one or two more words, whether correct or not, from at least half of the subjects. The final experimental task was the spatial memory test, administered to all subjects except those in the Garden on the second day. Spatial memory could not be sensibly measured within the environment itself. The spatial memory test used the computer to present subjects with five different images of locations in the Garden and they were to click on a map of the Garden to guess the locations. To avoid cross-test cuing, the locations were never the same as the verbal task locations. The subjects began the test with three points for each slide for a total starting score of fifteen. The scoring worked deductively with one point deducted for each wrong guess, but with no negative scoring. Thus, no more than the three points could be lost per slide. Hint buttons were provided that subjects could click in order to view a video panning one direction or the other. This cost a point for each direction selected, but provided a great deal more locational information to help the subject. As a testament to the engaging nature of this test and subjects' motivation to perform well on it, it is worth noting that nearly all subjects continued to make guesses after losing all points for a slide, even when they were reminded that their scores would no longer be affected and that the experimenter was willing to provide the answer. All subjects completed a post-experimental questionnaire at the close of the experiment. The questionnaire asked various questions regarding the significance and subjective nature of the manipulations. It also included a question regarding the degree to which subjects anticipated the verbal memory test, with possible answers ranging from 0 to 6. Scores higher than 3 were grounds for excluding the entire data set because the memory of subjects providing such a rating could not have been incidental. After Contextual Cue Selection 46 completing the post-experimental questionnaire, subjects were debriefed and thanked for their participation. Results Power analyses were not conducted prior to either of the present experiments for three reasons: (1) Many of the considered factors and relations are either unique or derived from inconsistent previous findings, so there is not a straight-forward source for the estimation of the non-centrality parameter, (2) those factors for which power analyses are possible would almost certainly register as under-powered, but (3) given the great number and variety of measured factors and their interactions, there was a strong likelihood of some effect being found somewhere despite the power levels, so the present experiments would have been conducted regardless the results of any power analyses. Given the number of dependent measures used in the same or similar situations in these experiments, multivariate analyses of variance (MANOVAs) could have been used to allow comparative analyses of the relative contributions of each of the measures. M A N O V A s were in fact included in an earlier version of this dissertation, but the results were not deemed meaningfully interpretable to a degree that justified the added level of complication nor the degree of sophistication that had to be dropped from the individual A N O V A s in order to make them compatible for an overall M A N O V A . Furthermore, the only comparison between dependent measures of any real interest, the relation between mood and memory, is discernible in finer detail through the transformation of the affective measures into dichotomous independent variables for use in post hoc analyses of the sort to be described at the end of the results section. Cognitive Measures Verbal Memory. The analysis of variance ( A N O V A ) performed on the recall data found a significant main effect of the task variable, F(2,48) = 3.93, p < .05, which an Contex tua l C u e Selection 47 or thogonal contrast reduced to s ignif icant ly greater recal l o n the part of the integrated activities task subjects ( M = 18.72) over that of subjects i n bo th the independent ( M = 13.50) and redundant ( M = 14.28) activities task groups, F(l ,48) = 7.70, p < .01. The respective effect sizes of those differences, relative to the overa l l var iance i n the mode l , were .87 and .74 of a s tandard devia t ion. There was also a non-signif icant t rend t oward an interact ion be tween the reinstatement and task variables, F(2,48) = 2.85, p = .07. A l l of this can be seen i n F igure 6. There were no effects of f ie ld- independence u p o n verbal memory , nor d i d i t interact w i t h the task or reinstatement factors. 30 T 25-j -"« 2 0 -J -G rt 15-j m _ -10-j 5-j 0 " I X Reinstatement Cl None |U Physical Independent Activities Redundant Activities Integrated Activities T y p e of Task F i g u r e 6. Reca l l b y levels of task and context reinstatement, Expe r imen t 1. Spat ia l M e m o r y . There were no significant effects of the task or reinstatement variables u p o n subjects' m e m o r y for locations i n the Garden . The means, out of a total possible of 15, were 3.00 for bo th the independent and redundant activit ies task groups, and 4.44 for the integrated activities task group. A d d i n g sex as a factor i n the analysis p roduced no change. There were no m a i n effects of sex o n spat ial m e m o r y , nor d i d it interact w i t h either the task or the reinstatement factors. Contextual Cue Selection 48 Affective Measures Mood and Environment Ratings. The fact that half the subjects were in a different environment on the second day necessitated a dual approach to the mood data analyses. A n A N O V A was conducted for the data from each of the two days. The data for the first day were analyzed using a between-within design with task as the between-subjects variable and time as the within-subjects variable. The three levels of task (independent redundant, and integrated activities) have already been described and time had two levels, the baseline and post-task ratings. The data from the second day were analyzed using a two-way between-groups design with task and reinstatement as the independent variables. Furthermore, separate analyses were conducted for each of the four dimensions into which the two mood measures can be divided, positive affect, negative affect, pleasantness, and arousal. Table 1 lists all four mean mood ratings by levels of the time and task variables. The means from the second day are not split by level of reinstatement because that variable had no effect, all main effect Fs(l,48) < 1 and all interactive Fs(2,48) < 1.5, ps > .25. Looking at the P A N A S data, the only finding of interest for the positive affect dimension was a trend toward increased positive affect from baseline to post-task on the first day, F(l,51) = 3.54, p = .07. For the negative affect dimension, time also had an effect, showing a significant decrease from baseline to post-task, F(l,51) = 5.81, p < .02, with an effect size of .47 of a standard deviation. In addition, there was a significant difference in negative affect on the first day depending upon the type of task performed, F(2,51) = 3.45, p < .05. A Tukey test narrowed the effect to a significantly greater amount of negative affect for subjects in the integrated activities condition, as opposed to those in the independent activities condition (p < .05), an effect with a size of .61 of a standard deviation. There was a similar, though non-significant, trend in the data for the second day, F(2,48) = 2.49, p = .09. This was the only effect on the second day approaching statistical significance. Contextual Cue Selection 49 Task Baseline T i m e Post-Task Day 2 Positive Affect Independent Redundant Integrated 28.78 28.98 31.72 30.61 29.11 28.28 Positive Affect 28.89 28.94 27.44 Positive Affect 29.28 31.17 28.28 Negative Affect Independent Redundant Integrated 11.22 12.67 11.00 11.61 10.83 12.06 Negative Affect 12.94 11.72 12.61 Negative Affect 13.83 12.11 12.72 Pleasant Independent Redundant Integrated 1.67 1.76 3.00 2.57 2.61 2.28 Pleasant 1.56 2.17 2.11 Pleasant 2.06 2.56 2.11 Arousal Independent Redundant Integrated .72 .87 .72 .81 1.28 1.11 Arousal .83 1.22 1.28 Arousal 1.06 .50 .78 Table 1. Mean ratings of mood in Experiment 1, by levels of task and time. Turning to the mood grid, the pleasantness dimension displayed the same time effect as the positive affect dimension, with a significant increase i n pleasantness from baseline to post-task, F(l,51) = 25.83, p < .0001, and an effect size of .98 of a standard deviation. There was also a trend towards an interaction of the task variable with time, F(2,51) - 2.66, p - .08. Finally, there were no significant effects upon the arousal dimension of mood in this experiment. As wi th the mood data, the environmental ratings had to be divided prior to analysis due to the different environments used on the second day of the experiment. In this case, the data were split by the levels of the reinstatement variable, none versus physical. The ratings from subjects returning to the Garden on the second day (physical reinstatement) could be analyzed using between-within A N O V A s nearly identical to the analyses conducted for the first day mood ratings. The only difference in this case was that the levels of the time variable were day 1 and day 2. In the case of ratings gathered Contextual Cue Selection 50 from subjects placed in different environments on the two days ("none" reinstatement), separate A N O V A s were conducted for ratings made on each of the experimental days. No effects were found for the complexity nor the preference ratings of the environment for either reinstatement group. The P A N A S and mood grid dimensions, however, were more sensitive to the experimental manipulations. With regard to the data from the physical reinstatement subjects, the time variable again showed the most consistent effect across all the affective dimensions except arousal. The positive affect of the environment was rated significantly lower upon returning the second day (M = 29.59) then it had been on the first day (M = 32.89), F(l,24) = 7.28, p < .02, an effect with a size of .73 of a standard deviation. The negative affect dimension also showed a decrease in affective intensity from the first day (M = 11.93) to the second (M = 10.89), F(l,24) = 5.08, p < .05, with an effect size of .62 of a standard deviation. There was no main effect of time on the rated pleasantness of the environment, but time significantly interacted with task, F(2,24) = 6.00, p < .01. A n orthogonal contrast found the change across time for the redundant activities group to be significantly different from that of the other two task groups, F(l,24) = 11.92, p < .005. The effect sizes were 1.90 and 2.09 standard deviations for the differences in change compared to the independent and integrative activities conditions respectively. Looking at Figure 7, it is possible to see that the change due to time was in fact reversed for the redundant activities group, compared to the others. There were no significant effects of the independent variables upon any of the affective environmental ratings from the none reinstatement subjects. Although mood was not explicitly manipulated, exploratory analyses examined the possible influence of subjects' moods upon memory performance in an attempt to find some indication of mood-dependent effects. The mood data were dichotomized in three separate fashions for these analyses and then compared to both the verbal recall and the spatial memory measures. The first dichotomy was between those subjects exhibiting a small versus a large difference in their mood ratings from the first to the second experimental day. This nominal indicator, referred to as mood difference, was Contex tua l C u e Selection 51 4-1 I 3 X I I T i m e • D a y l • Day 2 0 Independent Ac t i v i t i es Redundant Ac t i v i t i es Integrated Ac t i v i t i e s Type of Task Figure 7. E n v i r o n m e n t a l pleasantness b y levels of t ime and task, Expe r imen t 1. obtained b y calcula t ing the absolute value of the difference be tween the first and second day m o o d ratings a n d separating the result into large versus sma l l difference b y means of a m e d i a n split . There were no significant differences i n ve rba l or spatial m e m o r y i n Exper imen t 1 due to this cr i ter ion. The second and t h i rd dichotomies i n m o o d ratings relevant to the m e m o r y measures i n v o l v e d the moods occur r ing o n the first day, d u r i n g the learn ing tasks. These were d i v i d e d i n two ways , depend ing o n the type of measure. The i n d i v i d u a l P A N A S ratings of posi t ive and negative affect and the m o o d g r i d d imens ions of pleasantness and arousal were a l l subjected to a m e d i a n spl i t into h i g h and l o w ratings of each m o o d d imens ion , referred to as the va lenced intensity. The t w o m o o d g r i d dimensions , each h a v i n g posi t ive and negative valences, were also subjected to a m e d i a n spl i t based o n their absolute values, referred to as their absolute intensity (also d i v i d e d into h i g h and low) , regardless of valence. For the pleasantness d imens ion , however , the latter d i cho tomy made no difference compared to the p rev ious because Contextual Cue Selection 52 there were no subjects providing negative ratings on the pleasantness dimension of mood. The results indicated a marginally significant difference in recall based on the valenced intensity of negative affect on the first day, f(52) = 1.97, p = .054. The recall (M = 16.97) of subjects experiencing high negative affect was higher than that (M = 13.52) of those experiencing low negative affect. This effect did not interact with the type of task, F(2,48) = 1.13, p > .25. No other differences were found for any of the dichotomies of any of the other mood dimensions. D i s c u s s i o n Although there was not a preponderance of significant results in the present experiment, and no evidence of context-dependent memory effects (the possible reasons for which will be discussed in the General Discussion section), the different modes of processing the stimulus words did have an effect on later memory for those words, and there were some notable changes in the mood and environment ratings. C o g n i t i v e I m p l i c a t i o n s The realization of the predicted superiority of verbal memory on the part of integrated activities subjects supports the hypothesis that carefully patterned contextual integration can increase the memorability of individual contextual elements. Within the confines of this single experiment, however, there is no indication that the result was necessarily a function of the integrative nature of the task. Other possible explanations will be informed by a consideration of the affective results. The lack of any effects on spatial memory was not entirely unexpected. The number of subjects per group was particularly low for this measure because only the subjects returning to the laboratory on the second day could be tested. A replication at least of the previously found effect of sex (McGuiness & Sparks, 1983; Moffat et al., 1998) would have instilled greater Contextual Cue Selection 53 confidence in the validity of the new measure, but that was only a minor concern for the present project. Affective Implications The overall decrease in negative affect and increase in pleasantness from the beginning to the end of the first day confirm that the Nitobe Memorial Garden was generally considered a pleasant place to be, regardless of what task was performed there. The significantly higher overall negative affect for subjects performing the integrated activities task, however, suggests that this task may have been more annoying, anxiety-provoking, or otherwise unpleasant to subjects than the independent activities task. Given that even the baseline reading was higher, it may be that the instructions for the task alone were sufficient to induce a negative affective response. This makes intuitive sense since the integrated activities task was more complex than the others and greater complexity has been known to decrease affective appeal (Berlyne, 1963). Such a conclusion runs counter to the assumption that contextually integrated situations should be more desirable and easier to process. Furthermore, it provides an alternate explanation for the superior verbal memory of integrated activities subjects. The greater negative affect of subjects in this condition may have contributed to the memory improvement, for instance by rendering all associated material more memorable. Equally likely, the negative affect and greater recall performance may have both resulted from greater effort spent on the integrated activities task. Although the post hoc analysis of the effect of mood on memory indicated greater recall for those subjects experiencing greater negative affect the first day, this effect did not interact with the task variable, so there is no independent support of the foregoing argument. That is, greater negative affect during learning appears to have increased later recall regardless the method for processing the to-be-remembered stimuli. Since there was no change in recall based on the mood difference from the first Contextual Cue Selection 54 to the second day, mood congruence cannot adequately explain this result. Thus a slight degree of negative affect may serve to enhance memory for materials learned during that mood, regardless the other factors present. Considering the ratings of the environment, the fact that both positive and negative affective qualities of the environment were rated lower on the second day suggests that there may have been some degree of habituation to the environment generating less pronounced affective reactions upon remrning to the familiar place. This type of environmental habituation, and the detection of novelty against the backdrop of an otherwise familiar context, is considered by Nadel and Willner (1980) to underlie cognitive mapping processes. The significant interaction of task with time in the ratings of environmental pleasantness may seem to contradict this interpretation, given that the redundant activities group exhibited an opposite trend, attributing more pleasantness to the Garden on the second day than on the first (see Figure 7). Since that group's day 1 rating of the Garden occurred upon their first entering it, however, the reverse trend in their ratings may have been due to a lower initial rating resulting from the limited environmental information available. EXPERIMENT 2 The redundant activities task in Experiment 1 was derived from a theoretical stance internal to the concept of contextual integration. That is, given the ideal of full contextual integration, a task was created that would mimic its complexity without the integration. A primary goal of Experiment 2 was the isolation of the personal significance factor inherent in full contextual integration, with a means of doing so based upon existing theory and research. The provision versus restriction of word choice during the verbal task was used to systematically add or remove a person-task relation from the experimental context. This was done with the hope of creating or negating an effect like that exerted by the element of control (Baum & Valins, 1979; Contex tua l C u e Selection 55 Evans & C o h e n , 1987; Glass and Singer, 1972) f rom the env i ronmenta l literature, or l ike the generat ion effect (Slamecka & Graf, 1978) f rom the cogni t ive literature. In addi t ion , the resul t ing tasks, w h e n added to the independent and integrated activities tasks, p r o v i d e d a systematic progress ion th rough four levels of increas ing contextual integrat ion. B y a d d i n g the element of choice to the independent activities task, an act ivat ion of the person-task re la t ion was added to the person-environment relat ion, w i thou t a h igher- level re la t ion be tween the two. The resul t ing task is labeled "Choice O n l y " i n F igure 8. The other task created for this experiment s i m p l y r e m o v e d the element of choice f rom the integrated activities task. This e l iminated the person-task re la t ion f rom the n o r m a l l y wel l - integrated w o r d h i d i n g po r t i on of that task. W h a t r emained was a re la t ion be tween the person and the envi ronment that was then related to the task-envi ronment relat ion, but w i t h no significant connect ion be tween the person and the task. This task cond i t i on is referred to as " N o Choice" i n F igure 8. Level of Contextual Integraton (Task) Independent Activities Choice Only No Choice Integrated Activities n = 9 n = 9 n-9 n-9 None | n = 9 n = 9 n = 9 Yl — 9 | Mental F i g u r e 8. The des ign for Expe r imen t 2. The expected effect of these progress ively more integrated task modes was an incremental increase i n recal l performance a n d affective significance across the groups. That is, i n the case of recal l for instance, the independent and integrated activities task modes were expected to induce the lowest and highest levels of recal l , as demonstrated Contextual Cue Selection 5 6 in Experiment 1, with choice only and no choice falling between them in that order. If, on the other hand, contextual integration owes its influence entirely to the person-environment relation, perhaps through something like personal significance or control, then the two conditions involving word choice should influence a more pronounced, generally positive, affect and greater recall than the other two conditions. In addition, there should be no difference between the two choice nor between the two no-choice conditions. There were no specific hypotheses with regard to context-dependent recall, just the same general expectation, mentioned in relation to Experiment 1, that the mid-levels of contextual integration should be more susceptible than the extreme levels to variations in context across experimental sessions, hypotheses generated prior to the realization that Experiment 1 did not demonstrate context-dependent memory. Method Subjects The subjects in this experiment Were 72 men and women, ranging in age from 18 to 27 years old, randomly chosen from among those University of British Columbia psychology undergraduates responding to a departmental advertisement. All subjects received extra course credit for their participation. Sex and field-independence were balanced across groups as in Experiment 1. Materials and Equipment All materials and equipment in this experiment were identical to those used in Experiment 1. Design and Procedure The design of the experiment is illustrated in Figure 8. It is very similar to the design of Experiment 1. On the first day, the procedures for Experiment 2 were nearly indistinguishable from those of Experiment 1. The independent and integrated activities task procedures Contextual Cue Selection 57 were not only identical to those for Experiment 1, but thanks to the computer-controlled double-blind procedure, once both experiments were running concurrently it was impossible to tell which was being run when independent or integrated activities was the task. Furthermore, to the untrained eye, the two new tasks did not appear different from their predecessors, since the only change was the word-choice manipulation. O n the second day of the experiment, things began a little differently for half the subjects, those assigned to the mental reinstatement condition. Subjects in this condition stayed in the laboratory and were instructed to spend a few minutes thinking about the Garden as vividly as possible, including the entire path traveled. It was suggested that they might want to imagine not only the visual sensations experienced there, but the sounds and smells as well. They were also encouraged to close their eyes if doing so would help them mentally return to the Garden. The mental reinstatement condition was modeled after a similar condition used in Smith's (1979) research, and the instructions were nearly identical to his. After the mental reinstatement, or right away for subjects in the none condition of the reinstatement variable, subjects rated their mood and the environment, in a counter-balanced order across subjects. A n important detail about the environment ratings is that the mental reinstatement subjects were asked to rate the Garden environment they had just mentally revisited rather than their current laboratory surroundings. This further encouraged the reactivation of any associations with, or cues from, the previous day's environment. There was also a still image of the view from the previous day's rating location in the Garden displayed during the affective ratings. Results Cognitive Measures Verbal Memory. The A N O V A performed on the recall data found a significant main effect of the task variable, F(3,56) = 11.81, p < .0001, an effect that significantly Contextual Cue Selection 58 interacted with subjects' categorization as either field-dependent or field-independent (based upon a median-split of the project-wide CAB-Cf scores), F(3,56) = 3.12, p < .05. Orthogonal contrasts reduced the main effect of task to a significant difference between the recall performance of the two non-word-hiding groups, independent activities and choice only, and the performance of the two word-hiding groups, no choice and integrated activities, F(l,56) = 29.04, p < .0001. The size of this effect was 1.21 standard deviations. In addition, contrasts found the interaction between task and C A B - C f group (independent versus dependent) to lie primarily in greater memory performance for dependent subjects in the choice only group, when compared to that of dependent subjects in the independent activities group, F(l,56) = 10.55, p < .002 (effect size = 1.53 standard deviations), with no corresponding difference for the field-independent subjects in the same groups, F(l,56) < 1. These results are displayed in Figure 9. Spatial Memory. After reductions in the model to eliminate such non-significant factors as reinstatement and the interaction between C A B - C f group and sex, a main effect was found of sex on spatial memory, F(l,63) = 4.60, p < .05, indicating greater memory on the part of males (M = 5.25) versus females (M = 4.05), effect size = .46 of a standard deviation. There was also a marginally significant interaction between the task and C A B - C f group variables, F(3,63) = 2.65, p = .056. Figure 10 displays those data. Affective Measures Mood and Environment Ratings. The same dual approach used for the analyses of the mood data from Experiment 1 was used for Experiment 2. Table 2 lists all four mean mood ratings by levels of the time and task variables. Considering the P A N A S data from the first day, an A N O V A indicated a significant effect of time for both positive affect, F(l,68) = 4.65, p < .05, and negative affect, F(l,68) = 8.88, p < .005 with effect sizes of .34 and .50, respectively. Positive affect increased while negative affect decreased from baseline to post-task. There were no significant results in the day 2 positive or negative affect ratings. Contextual Cue Selection 59 30' 25 H - 20 H rt u C rt <u 15 H 10 H 0 1 X X _I_ X Independent Choice Only No Choice Integrated Activities Activities C A B - C f Group I"! Dependent | j | Independent Type of Task Figure 9. Recall by levels of task and CAB-Cf group, Experiment 2. 15-1 g 10' 5H JL JL X X X X C A B - C f Group n Dependent 111 Independent 0 1 I ' I I *•• I » Independent Choice Only No Choice Integrated Activities Activities Type of Task Figure 10. Spatial memory by levels of task, Experiment Contex tua l C u e Selection 60 T i m e Task Base l i ne Post -Task D a y 2 itive : ect I ndependen t 29.11 31.06 26.28 itive : ect C h o i c e O n l y 29.44 29.47 31.56 31.32 28.17 26.88 cn >+H n ° < N o C h o i c e 28.28 29.56 26.39 Integrated 31.06 33.11 26.67 cu Independen t 13.06 12.61 12.39 ativ feet C h o i c e O n l y 13.78 13.31 10.78 12.06 13.28 12.31 CD <r N o C h o i c e 13.06 11.83 11.78 Integrated 13.33 13.00 11.78 ntness I ndependen t 1.89 2.17 1.67 ntness C h o i c e O n l y 1.44 1.56 2.72 2.47 1.67 1.72 Pleasa N o C h o i c e 1.61 2.78 1.67 Pleasa Integrated 1.28 2.22 1.89 Independen t .56 1.22 .22 C h o i c e O n l y 1.28 1.04 .78 .76 1.50 .89 Aro N o C h o i c e 1.00 .72 1.44 Integrated 1.33 .33 .39 T a b l e 2. M e a n ratings of m o o d i n Exper iment 2, b y levels of task a n d t ime. A N O V A s conducted o n the m o o d g r i d data found two significant effects, one for the pleasantness d i m e n s i o n o n the first day and one for the arousal d i m e n s i o n o n the second day. The results indicate that o n the first day pleasantness signif icantly increased f rom baseline to post-task, F(l ,68) = 25.29, p < .0001, w i t h an effect size of .83, and o n the second day arousal leve l was signif icant ly dependent u p o n the type of task per formed o n the first day, F(3,64) = 3.61, p < .02. A n or thogonal contrast f ound the latter result attributable to the greater arousal experienced b y the choice o n l y a n d no choice groups compared to the other two, F(l ,64) = 10.72, p < .002, w i t h an effect size of .77. Contextual Cue Selection 61 The environmental rating analyses were divided as in Experiment 1 to insure that all subjects in each analysis were rating the same environment. Once again, no effects were found for the complexity nor the preference ratings of the environment for either reinstatement group. Also similar to Experiment 1, there were no significant effects involving any of the environmental ratings made by the none reinstatement subjects. For the mental reinstatement data, positive affect ratings of the environment decreased significantly from the first (M = 32.89) to the second day (M = 30.03), F(l,32) = 12.15, p < .002, a change with an effect size of .82. Ratings of the negative affect of the environment also decreased from day 1 (M = 12.08) to day 2 (M = 10.83), F(l,32) = 12.02, p < .002, with an effect size of .82. Finally, the arousal ratings of the environment showed a marginally significant increase from the first (M = .53) to the second day (M = 1.00), F(l,32) - 4.11, v = -051, an effect with a size of .47. As with Experiment 1, there were analyses conducted upon the present data to determine the influence of mood on memory performance. Considering the same three dichotomies of the mood data mentioned with regard to the results of Experiment 1, across the four different dimensions, the only significant difference was between the spatial memory (M = 5.2) of subjects experiencing a large mood difference from the first to the second day and that (M = 3.9) of those experiencing a small mood difference, £(70) = 2.01, p < .05. This effect was not found to interact with the type of task used during the first day, F(l,64) < 1. It may, however, be a by-product of the significant difference in the first day's positive affect between the two groups (large difference M = 33.92, small difference M = 28.57), f(70) = 3.65, p < .0005. Discussion Cognitive Implications Matching the results from Experiment 1, Experiment 2 also failed to find a context-dependent memory effect. The main effect of task on verbal recall and its interaction with the cognitive style of the subjects (CAB-Cf group) is, however, rather Contextual Cue Selection 62 informative and generally supports the contextual integration approach. By showing improved memory for subjects whose task connects verbal stimuli to the environment, regardless whether the stimuli are chosen or not, the main effect of task supports the hypothesis that the level of contextual integration has more of an effect upon recall than does the element of choice alone. The interaction with the CAB-Cf group factor suggests that this may only be the case for field-independent individuals. Field-dependent subjects show the same general trend in memory performance, but if the task-structuring influence of the word hiding task is absent, as in the independent activities and choice only groups, the subjects become dependent upon the ability to choose their words as a means of reducing the chaotic nature of the task. Field-independent subjects are probably not affected in this way because they are better able to generate their own structure for the less integrative tasks, making the element of choice less important as a guide for reducing the ambiguity of the situation. With regard to the spatial memory data, it was encouraging to obtain results commensurate with the literature this time around. By replicating past findings indicating better spatial ability for men (McGuiness & Sparks, 1983; Moffat, Hampson, & Hatzipantelis, 1998), the unique measure of spatial memory used in this experiment satisfied one of the most important criteria for establishing the validity of a new empirical tool. It is also interesting to note that the spatial memory data follow a trend similar to that of the recall data, with the exception that the opposite groups appear to have improved memory performance. The result was not quite significant, but the possibility that increased contextual integration with specific locations might detract from attention to the layout of the overall environment is intriguing. If the trend is indicative of an actual tendency, then this would somewhat compromise the assumption that contextual integration leads to a better overall internal representation of the entire situation. On the other hand, it is important to bear in mind that the spatial memory was global locational memory and the locations tested were never the same locations where the verbal tasks were performed. The testing method may also have Contextual Cue Selection 63 been non-optimal because it tested map-like locational knowledge when what was learned by the subjects was the knowledge gained by actively exploring an environment. Considering evidence that these are distinctly different forms of spatial knowledge (see Schacter & Nadel, 1991, for a review), there may have been an insufficient transfer of knowledge to allow such a mismatch between learning and testing methods. Many subjects in fact commented, while attempting the spatial memory test, that they had focused more on landmarks and had not formed a useful impression of the layout of the Garden the previous day. It is also interesting to note that the spatial memory data for the independent and integrated activities task groups follow a pattern identical to that of the data from the "explore" and "spy" groups, respectively, in the Ward et al. (1988) experiment. They were testing memory for features of the environment rather than layout, but the precedent is striking nonetheless. Ward et al. (1988) explain their result in terms of the "narrower focus" of subjects in the "spy" condition (p. 4). Affective Implications The mood and environment rating results of this experiment in large part replicated those of Experiment 1. For the mood data, there were two exceptions. The greater overall negative affect reported by subjects in the integrated activities task condition in Experiment 1 was not replicated and there was a new effect in this experiment showing that arousal on day 2 was dependent upon the task performed. The fact that day 2 arousal was significantly higher for the two new groups, and not the independent nor the integrated activities groups, suggests that the intermediate levels of contextual integration comprising the tasks for the new groups may be associated with greater apprehension, anxiety, excitement, or some other such reaction related to high arousal. It is unusual that any such influence would function on the second day and not the first, but it is possible that the uncertain nature of the procedures on the second day created more anticipation, curiosity, or anxiety in subjects in the choice only and no Contextual Cue Selection 64 choice groups than in the other groups. According to the contextual integration approach, this would have resulted from the more disjointed nature of the choice only and no choice tasks, which would have rendered them less meaningful and therefore more subject to speculation regarding their purpose, hence the greater arousal. This fits with how Berlyne (1963) would describe the reaction to an incongruous stimulus. In the case of the environment ratings, the nearly identical decreases in affective quality, regardless of valence, might lend further support to the notion that subjects became habituated to the Garden environment. It is important to note, however, that in the present case the day 2 ratings were actually made based on memories of the environment considered while subjects were in the laboratory. Depending upon how likely it is that a person can become habituated to a remembered stimulus, it may be necessary to posit some other explanation of the attenuated day 2 environment ratings, such as regression to the mean. Finally, with regard to the effects of mood on memory, given that significantly greater positive affect on day 1 may account for the greater decrease in positive affect for the large versus the small difference subjects (although mood difference was a non-valenced dichotomy, nearly all changes in positive affect were in fact decreases), it is possible that it was the greater initial positive affect that determined a greater attention to environmental details, hence greater spatial memory on the part of large difference subjects. One disadvantage to this explanation is that it ignores the lack of a difference in spatial memory based upon the valenced intensity dichotomy of positive affect on the first day. That is, there is no greater spatial memory for subjects experiencing high versus low rated positive affect during initial exposure to the environment. A better explanation for the greater spatial memory of subjects exhibiting large differences in positive affect is that the decreased affect reflects the use of more elaborative processing. Without some type of thinking or elaboration there would be nothing to cause a change in affect. Thus greater affective change could reflect a greater degree of cognitive elaboration or further processing. In sum, it may be the case that the large Contextual Cue Selection 65 differences subjects were those who spent more cognitive resources processing the environment, which, in turn, led to increased memory for the environment as well as the change in affect over time. GENERAL DISCUSSION Despite the lack of contextual memory effects, the manipulations for the present experiments had a measurable influence on memory and affective ratings that revealed some interesting relations among the key contextual factors and other findings that are worthy of further discussion. There were various effects of the task variable in these experiments, suggesting that at least some of the manipulations were successful in their attempt to establish unique contextual meanings for the subjects. The fact that only one of those effects involved the mood measures suggests that the environment may have been a superior determinant of mood and/or the tasks were not particularly affectively loaded. The higher negative affect for the mid-level contextually integrated groups in Experiment 2, however, indicates that there may have been some reaction like greater apprehension, possibly due to the perceived arbitrary nature or incongruity of those tasks. This can be interpreted as support for the notion that incomplete contextual integration leads to more confusing perceptual phenomena. It is difficult to venture a guess as to why changes in the environment from the first to the second day did not affect recall in these experiments. Among the possible reasons are the incidental nature of the memory tasks, the moderate level of the moods experienced, and the rich and memorable nature of the environment. Given that most contextual memory research involves intentional forms of memory (Smith & Vela, 2001) , it is possible that the incidental memory used in the present experiments did not properly replicate the earlier studies. Perhaps intentional memory methods are required to ensure that the relevance of contextual factors will be Contextual Cue Selection 66 considered by subjects learning new material. A n d yet, incidental memory has been shown to benefit from other mnemonic techniques (Bellezza & Reddy, 1978). As for the moderate moods reported in the present experiments, given Eich's (1995) mood-mediation hypothesis, there may not have been any moods acquired that were distinctive enough to serve as memory cues. Without mood-mediation, the highly memorable nature of the Nitobe Garden may have maximized the usefulness of environmental cues for all groups. Neither of these speculations is particularly plausible, however, since the Nitobe Garden is the exact same environment used by Eich (1995) and thus at least some of the same effects should have been found. The effects of mood valence (Experiment 1) and change over time (Experiment 2) upon memory, while not interacting with the type of task, also attest to at least some influence of affective states upon both verbal and spatial memory. Perhaps the only explanation for the lack of context-dependent memory effects is that all of the experimental tasks and mood changes were equivalent in their disruption of any potential connections between the word lists and environmental cues. The task-related spatial memory results, though non-significant, suggest that there may have been more or less of a division of attention based on the type of task performed. That is, the tasks involving hiding spots, as Ward et al. (1988) suggest, may have induced too narrow a focus on local environmental features, adversely affecting overall spatial memory (or at least not augmenting it) but assisting verbal memory. The more purely exploratory tasks, on the other hand, did not re-direct attention in this way, leading to the opposite results. Again, the effect of the type of task upon the spatial memory results was non-significant in the present experiments, but the foregoing interpretation is somewhat supported by the frequent comments of subjects in the hiding conditions that the verbal task had prevented them from observing more of the Garden. This underscores the fact that, despite the potential for contextual integration to greatly enrich an experience, it may simply represent a complex Contextual Cue Selection 67 distraction if the integration is insufficient to engender a sense of meaningfulness, or if there is no obvious purpose to the integration. The interaction of field-independence with the independent activities and choice only tasks in Experiment 2 supports the assumption that a person's individual perspective, and perhaps other factors of personal relevance, play as much of a role in the cognitive ordering and representation of contextual elements as those elements themselves. It also supports the notion of a hierarchical nature to the cognitive ordering process, as discussed in the Introduction. Without the task structure offered by the word hiding procedure in the no choice and integrated activities task conditions, field-dependent subjects in the other two groups in Experiment 2 remembered more if they got to choose their words. The exact mechanisms by which this might have worked are speculative because field-dependence itself is not thoroughly understood. Given the general perception that field-dependent individuals have difficulty developing their own structure for cognitive tasks (Davis & Frank, 1979), the word-choice procedure in the present experiments may have filled a void for them. From this perspective, the present results support the general idea that field-dependent individuals rely more upon externally-imposed organizational approaches to information processing, whereas field-independent individuals are more likely to provide their own organization. Relevant conclusions in the previous literature include the observations that "field-dependent people are able to sample fully from sets of cues if the cues are provided in discrete form" (Goodenough, 1976, p. 678), and "the chances of field-independent individuals displaying superior recall increases as the inherent organization of the task material decreases." (Davis & Frank, 1979, p. 477) More generally, the field-independence finding is also important because it shows that it is possible to empirically create a threshold level of contextual structure and to dichotomize performance on either side of that threshold based on minor personal (field-independence) or procedural (element of choice) differences. Contextual Cue Selection 68 The most interesting result from the mood and environment data was the tendency for subjects to rate the same environment with a lower degree of affect upon a second encounter. Habituation and regression to the mean were already mentioned as possible explanations of this effect, but there is another. Given that people have been found to prefer environments with some degree of mystery (Kaplan, 1987), it is possible that affective ratings in the present experiments changed on the second day because the Garden no longer afforded any mystery after being explored the previous day. That would not, however, explain why ratings of negative affect decreased as well as those of positive affect. Perhaps the former decrease was due to some other factor, such as the reduction in arousal and associated negative sensations that Berlyne (1963) says follow a reduction in the incongruity or complexity of perceived stimulation. If both the foregoing speculations were indeed true, that would imply that familiarity can decrease both positive and negative affective qualities of a situation. There was not, however, the universal reduction in arousal in these experiments that such an interpretation would require. One over-arching conclusion that can be made regarding the present experiments is that multi-faceted approaches to research in psychology can be usefully organized using the guideline of contextual integration. Regardless of whether or not the human mind naturally utilizes such an approach, organizing the contextual elements of an experiment according to the person, environment, and task categories can help clarify the likely results of their interactions. Furthermore, assuring activation of all of the contextual categories and their relations greatly enhances one's chances of uncovering previously undetected interactions among the determinants of behavior. Contextual integration should not necessarily be limited to the three key factors, and their interrelations, as defined here. The person, task, and environment factors were chosen because they do a rather thorough job of circumscribing everything that comes together to create an environmental milieu. There are bound to be constraining factors that might replace one of the key factors defined here with some element more Contextual Cue Selection 69 relevant to a particular situation. At the very least it is possible that the key factors and their relations may have different positions in each individual's hierarchy of personal significance. A n d yet the factors settled on here are so all-encompassing that both of the qualifications just mentioned can be reinterpreted as changes in those very factors themselves. From this perspective it is difficult to imagine what other contributory factors there could be that do not fit under the umbrella of contextual integration. The possibility remains that contextual integration is just another way of saying something else, like depth of processing. Depth of processing remains the single most powerful remaining alternative explanation to what has been hypothesized here as the most basic way of looking at the psychological determinants of such key processes as attention and decision making. If thinking about words long enough to choose from among them qualifies as deep processing, then Experiment 2 may have already demonstrated the superior explanatory weight of contextual integration by showing that deep processing did not supercede the influence of contextual integration. The possibility still remains that they are part of the same concept or two parts to a still higher construct. The possible overlap of depth of processing with contextual integration is supported by the following quote: Finally, it seems necessary to bring in the principle of integration or congruity for a complete description of encoding. That is, memory performance is enhanced to the extent that the encoding question or context forms an integrated unit with the target word (Craik & Tulving, 1979, p. 291). This shows that at the very least contextual integration and depth of processing are using the same explanatory constructs and defining qualities. Further research will have to discern how much the concepts overlap. Contextual integration is not intended as another theoretical model so much as a general guideline for delineating the relevant constructs applicable to a given empirical question. Because it is modeled after the natural means by which the mind subdivides its surroundings, contextual integration might aid the rudimentary subdivision of an experimental context into its relevant and irrelevant parts. Not unlike the C o n t e x t u a l C u e Selection 70 e n v i r o n m e n t a l preference studies m e n t i o n e d i n the Introduct ion, it s h o u l d be possible to predict behaviors such as increased approach or avoidance to a g i v e n s t i m u l u s predicated u p o n , a m o n g other things, the m a t c h of a person's cognit ive set to the avai lable e n v i r o n m e n t a l affordances. B u t these are a l l h i g h a n d m i g h t y goals c o m p a r e d to the present status of this construct. F o r the m o m e n t , further research needs to more direct ly determine i f the contextual integrat ion a p p r o a c h is any m o r e capable of p r e d i c t i n g the patterns of h u m a n reactions to e n v i r o n m e n t a l s t i m u l a t i o n t h a n was i n d i c a t e d b y the current project. F u t u r e direct ions of the present research approach i n c l u d e the use of " v i r t u a l " environments , the use of a b r o a d var iety of env ironments , a n d p o s s i b l y the creation of tasks capable of m o r e c o m p l e x m a n i p u l a t i o n s of contextual relations. The use of a v i r t u a l e n v i r o n m e n t has already reached the p i l o t i n g phase a n d promises to a l l o w the other enhancements of m o r e var iety a n d m o r e complexity . F o r instance, the computer s i m u l a t i o n u s e d i n brief for these experiments is presently capable of d i s p l a y i n g either the N i t o b e G a r d e n or a n i n d o o r e n v i r o n m e n t a n d has the potent ia l to replace the verba l s t i m u l i w i t h pictures. Research i n v i r t u a l environments w i l l never replace the use of true p h y s i c a l s u r r o u n d i n g s , b u t the rea l i sm of the computer s imulat ions is a lways i m p r o v i n g a n d the greater sophist icat ion of the person-environment relations that can be i n s t i l l e d u s i n g t h e m is rather persuasive. A d d to these considerations the potential to post such a s i m u l a t i o n , w i t h f u l l data collecting capabilities, o n the internet a n d s u d d e n l y the possibi l i t ies appear t r u l y endless. W h e t h e r i n v i r t u a l or actual environments , the current project demonstrates the importance of c o n s i d e r i n g a f u l l s p e c t r u m of contextual factors w h e n d e s i g n i n g b e h a v i o r a l research, a n d the potent ia l to d o so w i t h o u t sacri f ic ing m u c h i n the w a y of exper imenta l rigor. Contextual Cue Selection 71 REFERENCES Altman, I., & Rogoff, B. (1987). Chapter 1: World views in psychology: Trait, Interactional, Organismic, and Transactional perspectives. In D. Stokols & I. Altman (Eds.), Handbook of environmental psychology (Vol. 1, pp. 7-40). New York: John Wiley & Sons. Anderson, J. R. (1990). The adaptive character of thought. Hillsdale, NJ: Lawrence Erlbaum. Anderson, J. R., & Bower, G. H . (1973). Human Associative Memory. Washington, D C : Winston & Sons. Anderson, J. R., & Schooler, L. J. (1991). Reflections of the environment in memory. Psychological Science. 2, 396-408. Appleton, J. (1984). Prospect and refuge re-visited. Landscape Journal, 3, 91-103. Barker, R. G. (1968). Ecological psychology: Concepts and methods for studying the environment of human behavior. Stanford, C A : Stanford University Press. Barker, R. G. (1987). Chapter 40: Prospecting in Environmental Psychology: Oskaloosa revisited. In D. Stokols & I. Altman (Eds.), Handbook of environmental psychology (Vol. 2, pp. 1413-1432). New York: John Wiley & Sons. Baum, A. , & Valins, S. (1979). Architectural mediation of residential density and control: Crowding and the regulation of social contact. Advances in Experimental Social Psychology. 12,131-175. Bellezza, F. S., & Reddy, B. G. (1978). Mnemonic devices and natural memory. Bulletin of the Psychonomic Society. 11, 277-280. Berlyne, D. E. (1960). Conflict, arousal, and curiosity. New York: McGraw-Hill. Berlyne, D. E. (1963). Complexity and incongruity variables as determinants of exploratory choice and evaluative ratings. Canadian Journal of Psychology, 17, 274-290. Blaney, P. H . (1986). Affect and memory: A review. Psychological Bulletin. 99, 229-246. Bower, G. H . (1981). Mood and Memory. American Psychologist, 36, 129-148. Bower, G. H . , & Mayer, J. D. (1989). In search of mood-dependent retrieval. Journal of Social Behavior and Personality, 4,133-168. Cermak, L. S., & Craik, F. I. M . (Eds.). (1979). Levels of processing in human memory. Hillsdale, NJ: Lawrence Erlbaum. Collins, A . M . , & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82, 407-428. C o n t e x t u a l C u e Selection 72 C o r e n , S., Porac, C , & Theodor, L . H . (1986). The effects of perceptual set o n the shape a n d apparent d e p t h of subjective contours. Percept ion & Psychophys ics , 39, 327-333. C r a i k , F. I. M . , & T u l v i n g , E. (1975). D e p t h of processing a n d the retention of w o r d s i n ep isodic m e m o r y . J o u r n a l of E x p e r i m e n t a l Psychology: Genera l , 104, 268-294. D a v i s , J . K. , & F r a n k , B. M . (1979). L e a r n i n g a n d m e m o r y of f ie ld independent-dependent i n d i v i d u a l s . Journal of Research i n Personality , 13, 469-479. E i c h , J . E . (1980). The cue-dependent nature of state-dependent retr ieval . M e m o r y & C o g n i t i o n , 8,157-173. E i c h , E . (1985). Context , memory , a n d integrated i tem/context imagery. Tournal of E x p e r i m e n t a l P s y c h o l o g y : L e a r n i n g , M e m o r y , a n d C o g n i t i o n , H , 764-770. E i c h , E . (1995). M o o d as a mediator of place dependent m e m o r y . J o u r n a l of E x p e r i m e n t a l Psychology: Genera l , 124, 293-308. E i c h , J . E. , Weingartner, H . , S t i l l m a n , R. C , & G i l l i n , J . C . (1975). State-dependent accessibil ity of retr ieval cues i n the retention of a categorized list. J o u r n a l of V e r b a l L e a r n i n g a n d V e r b a l Behavior , 14, 408-417. E i c h , E. , & B i r n b a u m , I. M . (1982). Repet i t ion, cu ing , a n d state-dependent memory. M e m o r y & C o g n i t i o n . 10,103-114. E i c h , E. , & Metcal fe , J . (1989). M o o d dependent m e m o r y for internal versus external events. Tournal of E x p e r i m e n t a l Psychology: L e a r n i n g , M e m o r y , a n d C o g n i t i o n , 15, 443-455. Evans, G . W. , & C o h e n , S. (1987). Chapter 15: E n v i r o n m e n t a l stress. In D . Stokols & I. A l t m a n (Eds.), H a n d b o o k of e n v i r o n m e n t a l p s y c h o l o g y (Vol . 1, p p . 571-610). N e w York: J o h n W i l e y & Sons. Fernandez , A . , & G l e n b e r g , A . M . (1985). C h a n g i n g e n v i r o n m e n t a l context does not re l iab ly affect m e m o r y . M e m o r y & C o g n i t i o n , 13, 333-345. Forgas, J . P. (1999). N e t w o r k theories a n d b e y o n d . In T. D a l g l e i s h & M . J . P o w e r (Eds.), H a n d b o o k of c o g n i t i o n a n d e m o t i o n (pp. 591-611). Chichester: J o h n W i l e y & Sons. Genereux, R. L . , W a r d , L . M . , & Russel l , J . A . (1983). The b e h a v i o r a l component i n the m e a n i n g of places. J o u r n a l of E n v i r o n m e n t a l Psychology , 3, 43-55. G i b s o n , J . J. (1979). The ecological a p p r o a c h to v i s u a l percept ion. Boston: H o u g h t o n M i f f l i n . Glass, D. , & Singer, J . (1972). U r b a n Stress. N e w York: A c a d e m i c Press. G o d d e n , D . R., & Baddeley, A . D . (1975). Context-dependent m e m o r y i n t w o n a t u r a l environments : O n l a n d a n d underwater. B r i t i s h Tournal of P s y c h o l o g y , 66, 325-331. Contextual Cue Selection 73 Goodenough, D. R. (1976). The role of individual differences in field dependence as a factor in learning and memory. Psychological Bulletin, 83, 675-694. Hakstian, A . R., & Cattell, R. B. (1975). The Comprehensive Ability Battery (CAB). Champagne, IL: Institute for Personality and Ability Testing. Herzog, T. R., & Flynn-Smith, J. A . (2001). Preference and perceived danger as a function of the perceived curvature, length, and width of urban alleys. Environment and Behavior, 33, 653-666. Herzog, T. R., Kaplan, S., & Kaplan, R. (1976). The prediction of preference for familiar urban places. Environment and Behavior, 8, 627-645. Herzog, T. R., & Kutzli, G . E. (2002). Preference and perceived danger in field/forest settings. Environment and Behavior, 34, 819-835. Herzog, T. R., & Miller, E. J. (1998). The role of mystery in perceived danger and environmental preference. Environment and Behavior, 30, 429-449. Herzog, T. R., & Shier, R. L. (2000). Complexity, age, and building preference. Environment and Behavior, 32, 557-575. Jacobs, W. J., Thomas, K. G. R, Laurance, H . E., & Nadel, L . (1998). Place learning in virtual space: II. Topographical relations as one dimension of stimulus control. Learning and Motivation, 29, 288-308. Jacoby, L . L. , & Craik, F. I. M . (1979). Effects of elaboration of processing at encoding and retrieval: Trace distinctiveness and recovery of initial context. In L. S. Cermak & F. I. M . Craik (Eds.), Levels of processing in human memory (pp. 1-21). Hillsdale, NJ: Lawrence Erlbaum. Jenkins, J. J. (1974). Remember that old theory of memory? Well, forget it! American Psychologist. 29, 785-795. Jenkins, J. J. (1979). Four points to Remember: A tetrahedral model of memory experiments. In L. S. Cermak & F. I. M . Craik (Eds.), Levels of processing in human memory (pp. 429-446). Hillsdale, NJ: Lawrence Erlbaum. Jenkins, J. J. (1981). Can we have a fruitful cognitive psychology? In J. H . Flowers (Ed.), Nebraska symposium on motivation. Vol. 28: Cognitive processes (pp. 211-238). Lincoln: University of Nebraska Press. Kaplan, S. (1975). A n informal model for the prediction of preference. In E. H . Zube, R. O. Brush, & J. Gy. Fabos (Eds.), Landscape assessment: Values, perceptions, and resources (pp. 92-101). Stroudsburg, PA: Dowden, Hutchinson & Ross. Kaplan, S. (1987). Aesthetics, affect, and cognition: Environmental preference from an evolutionary perspective. Environment and Behavior, 19, 3-32. Kaplan, S., & Kaplan, R. (1983). Cognition and environment: Functioning in an uncertain world. A n n Arbor, MI: Ulrich's Bookstore. Contex tua l C u e Selection 74 K a p l a n , R., K a p l a n , S., & Ryan , R. L . (1998). W i t h people i n m i n d : D e s i g n and management of everyday nature. Washington, D C : Is land Press. K a p l a n , S., K a p l a n , R., & Wendt , J. S. (1972). Rated preference a n d complex i ty for natura l and u rban v i s u a l material . Percept ion & Psychophysics , 12, 354-356. Lockhar t , R. S., & Cra ik , F. I. M . (1990). Levels of processing: A retrospective commenta ry o n a f ramework for m e m o r y research. C a n a d i a n Tournal of Psycho logy . 44, 87-112. M c G u i n e s s , D . , & Sparks, J. (1983). Cogn i t ive style and cognit ive maps: Sex differences i n representations of a famil iar terrain. Tournal of M e n t a l Imagery, 7, 91-100. Metcalfe, J. (1996). Metacogni t ive processes. In E . L . Bjork and & R. A . Bjork (Eds.), M e m o r y (pp. 381-407). San Diego : A c a d e m i c Press. Metcalfe E ich , J. (1982). A composi te holographic associative recal l m o d e l . Psychologica l Rev i ew , 89, 627-661. Moffat, S. D . , H a m p s o n , E . , & Hatzipantel is , M . (1998). N a v i g a t i o n i n a "vir tual" maze: Sex differences and correlat ion w i t h psychometr ic measures of spat ial abi l i ty i n humans . E v o l u t i o n & H u m a n Behavior , 19, 73-87. N a d e l , L . , & Willner, J. (1980). Context and condi t ion ing: A place for space. Phys io log ica l Psycho logy , 8, 218-228. N e i l l , W . T., Beck, J. L . , Bottalico, K . S., & M o l l o y , R. D . (1990). Effects of intent ional versus inc identa l learn ing o n expl ici t and impl i c i t tests of memory . Tournal of Expe r imen ta l Psychology: Learn ing , M e m o r y , and C o g n i t i o n , 16, 457-463. Peron, E . , Purce l l , A . T., Staats, H . , Falchero, S., & L a m b , R. J. (1998). M o d e l s of preference for ou tdoor scenes: Some empi r i ca l evidence. E n v i r o n m e n t and Behavior , 30, 282-305. Purcel l , A . T., L a m b , R. ]., M a i n a r d i Peron, E . , & Falchero, S. (1994). Preference or preferences for landscape? jou rna l of E n v i r o n m e n t a l Psycho logy , 14,195-209. Purcel l , T., Peron, E . , & Berto, R. (2001). W h y do preferences differ be tween scene types? E n v i r o n m e n t and Behavior , 33, 93-106. Russel l , J. A . (2003). Core affect and the psychologica l construct ion of emot ion . Psycholog ica l Rev iew, 110,145-172. Russel l , J. A . , & Lan ius , U . F. (1984). Adap ta t i on level and the affective appraisal of environments . Tournal of E n v i r o n m e n t a l Psychology , 4,119-135. Russel l , J. A . , Weiss , A . , & Mende l sohn , G . A . (1989). Affect g r id : A single- i tem scale of pleasure a n d arousal , j ou rna l of Personal i ty and Social Psychology , 57, 493-502. Schacter, D . L . (1999). The seven sins of memory : Insights f rom psycho logy and cognit ive neuroscience. A m e r i c a n Psychologist , 54,182-203. Contextual Cue Selection 75 Schacter, D. L., & Nadel, L. (1991). Varieties of spatial memory: A problem for cognitive neuroscience. In R. G. Lister & H . J. Weingartner (Eds.), Perspectives on cognitive neuroscience (pp. 165-185). New York: Oxford University Press. Slamecka, N . J., & Graf, P. (1978). The generation effect: Delineation of a phenomenon. Tournal of Experimental Psychology: Human Learning and Memory, 4, 592-604. Smith, S. M . (1979). Remembering in and out of context. Tournal of Experimental Psychology: Learning, Memory, and Cognition, 5, 460-471. Smith, S. M . (1984). A comparison of two techniques for reducing context-dependent forgetting. Memory & Cognition, 12, 477-482. Smith, S. M . (1988). Environmental context-dependent memory. In G. M . Davies & D. M . Thomson (Eds.), Memory in context: Context in memory (pp. 13-34). Chichester: John Wiley & Sons. Smith, S. M . , Glenberg, A. , & Bjork, R. A . (1978). Environmental context and human memory. Memory & Cognition, 6, 342-353. Smith, S. M . , & Vela, E. (2001). Environmental context-dependent memory: A review and meta-analysis. Psychonomic Bulletin & Review, 8, 203-220. Sternberg, R. J., & Grigorenko, E. L. (2001). A capsule history of theory and research on styles. In R. J. Sternberg & L. Zhang (Eds.), Perspectives on thinking, learning, and cognitive styles (pp. 1-21). Mahwah, NJ: Lawrence Erlbaum. Stokols, D. (1979). A congruence analysis of human stress. In I. G. Sarason & C. D. Speilberger (Eds.), Stress and anxiety (pp. 27-53). Washington, D C : Hemisphere. Stokols, D. (1987). Chapter 2: Conceptual strategies of environmental psychology. In D. Stokols & I. Altman (Eds.), Handbook of environmental psychology (Vol. 1, pp. 41-70). New York: John Wiley & Sons. Suedfeld, P. (1991). Groups in isolation and confinement: Environments and experiences. In A . A . Harrison, Y. A . Clearwater, & C. P. McKay (Eds.), From Antarctica to outer space: Life in isolation and confinement. New York: Springer-Verlag. Toglia, M . P., & Battig, W. F. (1978). Handbook of semantic word norms. Hillsdale, NJ: Lawrence Erlbaum. Tryon, W. W. (2002). Network models contribute to cognitive and social neuroscience. American Psychologist, 57, 728. Wanner, H . E. (1968). O n remembering, forgetting, and understanding sentences: A study of the deep structure hypothesis. Unpublished doctoral dissertation, Harvard University. Ward, L. M . , & Russell, J. A . (1981a). Cognitive set and the perception of place. Environment and Behavior, 13, 610-632. Contextual Cue Selection 76 Ward, L . M . , & Russell, J. A . (1981b). The psychological representation of molar physical environments. Tournal of Experimental Psychology: General, 110,121-152. Ward, L. M . , Snodgrass, J., Chew, B., Russell, J. A . (1988). The role of plans in cognitive and affective responses to places. Journal of Environmental Psychology, 8,1-8. Ward, L. M . , & Suedfeld, P. (1973). Human responses to highway noise. Environmental Research, 6, 306-326. Watson, D., Clark, L. A. , & Tellegen, A . (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Tournal of Personality and Social Psychology, 54,1063-1070. Wilson, E. O. (1975). Sociobiology: The new synthesis. Cambridge, M A : Harvard University Press. Witkin, H . A. , Oltman, P. K., Raskin, E. , & Karp, S. A . (1971). Manual for the Embedded Figures Tests. Palo Alto, C A : Consulting Psychologists. Wohlwill, J. F. (1968). Amount of stimulus exploration and preference as differential functions of stimulus complexity. Perception & Psychophysics, 4, 307-312. Contex tua l C u e Selection 77 A P P E N D I X The Word Lists and their Semantic Ratings Togl i a and Bat t ig (1978) Clusters used: 7 & 8 • N u m b e r of words : 60 LIST R E S T R I C T I O N S : M i n w o r d length: 4 M a x w o r d length: 8 C o n l i m i t e d to 6.3 p r io r to m a n u a l modif icat ions L i s t l L i s t 2 Lis t 3 Lis t 4 Lis t 5 Lis t 6 apple ki te cereal b louse pear t rol ley plane nai l garlic t r ipod jacket chalk belt rocket bas in card spike l umber s p o o n coffee tobacco racquet glass knife vest sp r ing s tocking tomato trailer h o o k dol lar necklace rudder badge bone walnu t cigar seaweed candy mattress g o l d meda l magnet baseball bandage beret toast cage crystal w i n e canteen cucumber macaron i orange sardine potato boat pliers b r o o m plate C o n : 5.795 Img: 5.665 Cat : 5.770 M n g : 4.152 F a m : 6.099 N o a : 3.613 Pis: 4.326 O V E R A L L : Con=6.003 C o n : 6.085 Img : 5.959 Cat : 5.868 M n g : 4.534 F a m : 6.182 N o a : 4.053 Pis: 4.657 Img=5.715 C o n : 6.071 Img : 5.605 Cat : 5.496 M n g : 4.410 F a m : 5.957 N o a : 3.724 Pis: 4.109 C o n : 5.987 Img: 5.648 Cat : 5.494 M n g : 4.289 F a m : 5.776 N o a : 3.568 Pis: 4.292 C o n : 6.033 Img : 5.833 Cat : 5.725 M n g : 4.499 F a m : 6.237 N o a : 3.661 Pis: 4.169 C o n : 5.925 Img : 5.725 Cat : 5.479 M n g : 4.320 F a m : 5.954 N o a : 3.595 Pis: 3.947 Cat=5.633 Mng=4.368 Fam=6.011 Noa=3.685 Pls-4.212 Note: The fu l l names for the semantic d imens ions are: Concreteness, Imageabil i ty, Categor izabi l i ty , Mearungfulness, Fami l ia r i ty , N u m b e r of At t r ibutes , and Pleasantness. Pleasantness is treated as a b ipolar rating, w i t h a ra t ing of 4.000 ind ica t ing neutrali ty. The range for a l l ratings is 1-7. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items