<%BANNER%>

Influence of Time Spent in an Environment and Encoding Strategies on the Environmental Context Change Effect


PAGE 1

INFLUENCE OF TIME SPENT IN AN ENVIRONMENT AND EN CODING STRATEGIES ON THE ENVIRONMENTAL CO NTEXT CHANGE EFFECT By KHANH N. NGHIEM A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Khanh N. Nghiem

PAGE 3

iii ACKNOWLEDGMENTS It has been a long journey and I would like to thank all those who have helped me along the way. I thank my advisor, Dr. Pete r Delaney, for his guida nce and providing me with the necessary motivation to see this to the end. I would also like to thank my committee members Dr. Ira Fischler and Dr. Scott Miller, for their valuable input, support, and patience throughout th e duration of this thesis. I am deeply indebted to Lauren Saunders for her hard work and dedi cation in managing the la b and assisting me. I gratefully thank Jason Bendezu and Manue l Lopez for helping me collect a much needed portion of the data on very short noti ce. Michael Kung helped me with the study materials and I thank him for his assistance. I also thank the follo wing people for helping me collect the pilot data and experimental da ta: Chris Cardani, Jaye Murray, Matt Smith, Stevie Fisher, and Alfredo Llor eda. I would like to thank Keith McGregor for being an unshakeable supporter. Dr. Carolyn Tucker has been a professiona l inspiration and I would like to thank her for her advice, unde rstanding, and patience when I needed time away from work to collect data, write, and think Last but not least, I would like to thank the people in my personal life who may not have directly contributed to the completion of this thesis, but nonetheless are important to me. I would like to thank my parents who have supported my decisions and have been inspirational; I appreciate the sacrifices they have made in order to provide my brother and me a better future and education. I woul d like to thank my brother, who is my best

PAGE 4

iv friend and staunchest supporter. I am gratef ul for my best friend, Azam Khalid, who has provided many needed laughs, warm encouragements, and valuable advice.

PAGE 5

v TABLE OF CONTENTS ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES............................................................................................................vii FIGURES........................................................................................................................ .viii ABSTRACT....................................................................................................................... ix CHAPTER 1 ENVIRONMENTA L CONTEXT CHANGE EFFECT..........................................1 Habituation.............................................................................................................10 Strategy..................................................................................................................13 2 EXPERIMENT......................................................................................................19 Time and Strategy Predictions...............................................................................19 Working Memory Predictions................................................................................21 Methods..................................................................................................................22 Participants........................................................................................................23 Materials...........................................................................................................23 Procedure..........................................................................................................24 Encoding.....................................................................................................24 Recall..........................................................................................................25 Post-recall measures....................................................................................25 Results................................................................................................................. ...26 Time, Strategy, a nd Environment Variables......................................................26 Working Memory...............................................................................................27 Retrospective Reports........................................................................................30 Post Hoc Analyses.............................................................................................31 3 GENERAL DISCUSSION...................................................................................34 Environment..........................................................................................................34 Encoding Strategy.................................................................................................35 Time......................................................................................................................36 Working Memory..................................................................................................36 Population Sample Problems................................................................................38

PAGE 6

vi General Conclusion...............................................................................................39 LIST OF REFERENCES..................................................................................................41 BIOGRAPHICAL SKETCH............................................................................................47

PAGE 7

LIST OF TABLES Table Page 1-1. Recall Mean Proportions and Standard De viations for Early, Late, Match, Mismatch, Shallow Encoding, and Deep Encoding Conditions .................................................27 1-2. Recall Mean Proportions and Standard Deviations for Introductory Psychology Course Students .........................................................................................................32 1-3. Recall Mean Proportions and Standard Deviations for Upper-Division Psychology Course Students .........................................................................................................33 vii

PAGE 8

FIGURES Figure Page 1-1. Triangle C Span Task to Assess Working Memory Span Example Slides. ..............26 1-2. Best-fitting Regression Lines for Working Memory and Proportion of Words Recalled in Each Environment for Shallow Encoding Condition .............................29 1-3. Best-fitting Regression Lines for Working Memory and Proportion of Words Recalled in Each Environment for Deep Encoding Condition ..................................30 viii

PAGE 9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INFLUENCE OF TIME SPENT IN AN ENVIRONMENT AND EN CODING STRATEGIES ON THE ENVIRONMENTAL CO NTEXT CHANGE EFFECT By Khanh N. Nghiem August 2006 Chair: Ira Fischler Cochair: Peter F. Delaney Major Department: Psychology The environmental context change effect is the finding that forgetting occurs when participants are tested in environments that do not match the encoding environment. Regardless of the high frequency of anecdotes that are reported, environmental context effects have been di fficult to replicate or are modest. The following study includes one experiment that ex amined two variables that have not been extensively explored: time spent in the initial environment prior to studying and initial encoding strategies. Participants initia l encoding strategy was fixed and some participants studied the words immediately or after a delay. They were then asked to recall the words in an environment that e ither matched the encoding environment or mismatched. Participants who were required to do a shallow encoding strategy were expected to have better recall when th e environment matches, replicating the ix

PAGE 10

environmental context change effect from past research, and to do better when studying the words after a delay than when words were studied immediately. After a delay, participants may be habituating to the envi ronment and environmental cues will not be encoded with the target items. Participants who were required to engage in a deep encoding strategy were expected to not di splay any forgetting due to mismatch of environments regardless of duration of time spent in an environment. Deep encoding strategies allow for better retrieval cues in which there are inter-item associative processing where one item primes the retrieva l of another item. Working memory was also measured as past studies have found that working memory is related to attentional control. A 2 Time of Encoding 2 Encodi ng Strategy x 2 Environment between-subjects ANOVA with match and mismatch environments for the first factor, early and later encoding times as the second factor, and encodi ng strategy as the third factor revealed no significant main effects for time or significant interactions. The main effect of strategy was significant; participants in the deep enc oding strategy have highe r recall than those in the shallow encoding strategy regardless of time or if the environment matches or mismatches. There were no significant correl ations in working memo ry with any of the variables of interest (time, strategy, environments). The results suggest that perhaps encoding strategy plays a much stronger role in environmental context change effect than was previously expected. However, probl ems with the population sample may have resulted in undetectable effects. x

PAGE 11

CHAPTER 1 ENVIRONMENTAL CONTEXT CHANGE EFFECT In the coming-of-age movie Cinema Paradiso a famous filmmaker returns to his small hometown after an absence of 30 years for his mentors funeral. After encountering his house and the dilapidated movi e theater that was on ce the towns soul, memories of his childhood with the man who inspired his dream to become a filmmaker and memories of his first love come drifting up. The film is dramatic ally poignant in that these memories are recalled with a haunting tin ge of regret, nostalgi a, and realization of the joy and complexity of life. Our everyday life plays against a bac kdrop of environmental and sensory information. It is not surprising that our memo ries are intrinsically tied to contextually rich information. These contextual cues have the power to ev oke vivid and intense memories. All of us have at one point or other experienced thisa song playing that triggers a memory of a time we heard the tune catching a whiff of a scent that reminds us of a particular person, or returning to a prev ious place in the past and finding formerly unavailable memories are dredged up. Ou r memories are rich with these sensory, contextual cues. The body of research on context is vast, ex amining such diverse contextual factors as mood (Eich, 1980; 1985), internal contex t (Eich, 1975; 1980; Weingartner, Adefis, & Eich, 1976), and environmental context (S mith, 1979; 1984; 1986; Smith, Glenberg, & Bjork, 1978). Experimental context change pa radigms have been used extensively to study context reinstatement effects (Krafka & Penrod, 1985; Smith, 1979; 1984; 1985a), 1

PAGE 12

2 type of study material (Vel as study as cited in Smith, 1988; Eich, 1985; Steuck & Levys study as cited in Smith, 1988), and type of environment (Godden & Baddeley, 1975; 1980; Saufley, Otaka, & Bavaresco, 1986; Smith, 1985a; 1985b). The environmental context change study paradigm can be divi ded into two categories: first-order and second-order experimental paradigms (B jork & Richardson-Klavehn, 1988; Smith & Vela, 2001). First order experime ntal paradigms involve participants studying material in one environment and then being tested in th e same environment (control condition) or a different environment (context change conditi on). Godden and Baddeleys (1975) classic study is an example of a first order paradigm that examined context change where participants studied on land or under water and testing was either in the same environment or the other environment (e.g. study on land, recall under wa ter). First order experimental paradigms provide a simple manipulation of context change and thus criticism has arisen that participants may be ab le to mentally reinstate their environment. Second order experimental paradigms involve participants studying at more than one environment and with more than one list; henc e it is more difficult for participants to mentally reinstate their environment (Smith, 1984; Smith & Vela, 2001). Environmental context examined in this e xperimental proposal can be specified in the following dimensions: external (vs. internal ), incidental (vs. delib erate), not focal (vs. focal), and general (vs. specific) (Smith, 1988). Ostensibly, the distinction between external and internal context may seem straightforward. However upon further examination, external and inte rnal context is more entwin ed than may be originally assumed. External context is represented internally while internal context may alter the perception of external context. However, for the sake of simplicity, researchers have

PAGE 13

3 assumed that external contex t includes the physi cal aspect (i.e. environment) while internal context includes any factors that can be represented within a person (i.e. mood, thoughts, emotions). Environmental context is also incidental and not focal in that cues that are not meaningfully related to the study itselfthat is, it is outside the attentional focus of the participantare not actively utiliz ed by the participant. Also, environmental context is general; not any one specific aspect of the environment is manipulated but all parts of the physical environment. Of particular interest to th is experimental paper is e nvironmental context change. We can all relate particular episodes in which forgetting occurred due to a change in location: leaving a room to retrieve so mething only to forget it and suddenly remembering again when returning to the orig inal room or returning to a childhood house and remembering certain incidents that were un likely to be retrieved before. Regardless of the high frequency of anecdotes that are reported, environmental context effects have been difficult to replicate (Eich, 1985; Fe rnandez & Glenberg, 1985; Saufley, Otaka, & Bavaresco, 1985) or produce small effects. Smith and Vela (2001) performed a metaanalysis examining the reliability of contex t change and conclude d that environmental context effects were modest but reliable. The effect seems to be influenced by several variables; of those identified by Smith and Vela were: degree of differences between environments examined, time interval be tween learning and recalling, presence or absence of the same experimenter, and type of material learned (i.e. associative or nonassociative). Not surprisingly, the more the encodi ng environment matches the retrieval environment, the smaller the magnitude of forgetting due to changing environments.

PAGE 14

4 Smith and colleagues (1984, 1988, 2001) have sugge sted that the greater the resemblance between the encoding and testing environm ents, the more likely participants will mentally reinstate the original environment. Hence, the context change effect is reduced. In additional, increasing the time between learning and recall will bolster the context change effect. Accessibility of the memory traces may be weakened and hence more dependency on cue reinstatement, as in matchi ng environments. This results in greater reliance on contextual cues and a larger context change effect is observed. The presence of the same experimenter during both encodi ng and testing phases of the experiment for the mismatch environment conditions would contaminate context cues for the new environment. The experimenter is viewed as part of the environment and in a sense, can serve as a retrieval cue. Th erefore, when the same experi menter is present in the new environment, this environment has both new and old aspects (never exposed and previously exposed environments). Even more detrimental to experi mental results would be if the same experimenter is also present during the testing phase. Participants spend a large amount of time during experiments inter acting with the experimenter. Therefore, the presence of the same experimenter after an environment change is a poor experimental control. Consistent with th is idea, Smith and Vela found that switching experimenters for mismatch environment conditi ons resulted in increased context change effects. Interestingly, the type of material studied can influence the magnitude of context change effects. Materials that elicit interitem associative proce ssing in which one item may potentially prime the other during retrie val were found to redu ce the context change effect. Presumably, this was due to less re liance on environmental cues due to better retrieval cues.

PAGE 15

5 Several hypotheses have been proposed to explain environmental context change effects. One of the first explanations posited that the context change effect is dependent on proactive interference. Cont ext change that reduces proactive interference to benefit memory can be illustrated by naturalistic evidence. Smith (1988) points out vacations often facilitate creat ivity and productivity by aiding the vacationer in forgetting worries and trite ideas (i.e. writers block, stuck-in-a-rut conditions). We often refer to vacations as escapes and by going to an unf amiliar, novel environm ent, interest is renewed. In problem solving studies, it has been suggested that incubation or setting aside the problem when stuck and then later attempting to solve the problem may be dependent on context change (Smith, 1995). Smith suggests that a context change reduces fixation on and accessibility to wrong so lutions. This idea could be observed by our tendency to take a break when a task becomes too frus trating or tedious. Although reduction of proactive interference may be a likely explanation for overall better performance due to a change in context, it do es not adequately explain the forgetting that does occur when changing context. Early st udies of environmental context suggested that a context change lead to a reduction in proactive interference, which resulted in better recall (Da llett & Wilcox, 1968). However, th ese studies included learning in multiple contexts, which is a different paradigm than the one used in this study. Studying lists of words in multiple contexts leads to better recall than studying in the same context because the environment acts as an organi zational cue (Smith, 1982; 1984). This finding is supported by classroom studies in which no environmental context change effect was observed because students usually studied in many locations outside the classroom (Abernethy, 1940; Chens study as cited in Smith, 1988). Reduction of proactive

PAGE 16

6 interference as a plausible e xplanation of context change effect has not been widely accepted. The context change effect refers to forgetting information due to context shifts whereas release from proactive interferen ce generally leads to better recall or less forgetting (Wickens, 1970). Physical disruption during the changing of contexts was also suggested as a possibility for the context change effect. Strand (1970) found that a disruption during encoding of words also display forget ting and posited physical disruption during changing environments as driving the context change effect. After the word presentation, participants were interrupted and asked to wa it in the hallway. The time interval during the wait was equal to the amount of time walk ing participants to a new environment for the mismatch conditions. Participants waiting in the hallways ( match conditions) were returned to their original envi ronments for recall. This e xplanation has been discredited by studies that found reliable context cha nge effects even after accounting for the physical disruption (Smith, 1979; Smith, Gle nberg, & Bjork, 1978). Subsequent studies of context change effects included a physical disruption for all conditions to control for this potential confounding. Tulvings (1973) encoding specificity prin ciple states that the match between encoding and retrieval episodes is most importa nt for facilitating succ essful recall. The more similar the two episodes are in terms of intrinsic and extrinsic components, the more likely it is that the memory trace is recalled. This encodi ng specificity principle forms the underlying basis for the similarity principle, which provi des a context change effect explanation. According to the similarity principle, the context change effect is due to the higher degree of difference between mismatched environments than when the

PAGE 17

7 environments are the same. Ideally, environm ents with the most differences on the most dimensions will result in a larger context cha nge effect. However, this hypothesis is too general: it does not specify what these intrinsi c and extrinsic cues are. It is difficult to make predictions, experimentally manipulate variables that will pr oduce the most context change, or control for variability be cause these cues are not defined. The integration hypothesis (Baddeley, 1982) stat es that in order for the context to influence memory the study material must be integrated with contextual cues during encoding. Therefore, recall of the study mate rial is highly dependent on these contextual cues. This hypothesis has recently fo und support by studies examining context integration (Earles, Smith, & Park, 1994; Ei ch, 1985; Park, Smith, Morrell, Puglisi, & Dudley, 1990; Smith, Park, Earles, Shaw, & Whiting, 1998). These studies concluded that environmental context cues were utilized only when participants were instructed to. Not surprisingly, a context ch ange effect was observed onl y when participants were instructed to integrat e the study materials with elements of the environment. However, the integration process does not predict a ny context change effect for incidental environmental context effects. Past studi es have found reliable effects (Smith, 1979, 1985; 1986) and therefore this hypothesis has been viewed as not a strong explanation of environmental context change effect. Eichs (1995) mood mediation hypothesis st ates that environmental context change can be explained by mood-dependent memo ry. Participants moods vary across environmental episodes and these moods become intrinsically tied to environmental cues. Therefore, during retrieval, participants can access mood dependent cues as retrieval cues to recall. Hence, when the learning and tes ting environments match, participants are able

PAGE 18

8 to access their mood-dependent memory. A cont ext change effect is observed when the environments do not match and participants cannot access their mood-dependent memory. However, mood mediation hypothesis ha s been unable to explain the effects of material type eliciting associative or non-associative processing, test type, etc. (Smith, 1995; Smith & Vela, 2001). Researchers have pointed out that mood-dependent memory may be a reflection of mental context (S mith, 1995). It is unlikely that changing locations could produce a large enough change in mood that would be robust; even strong effort to manipulate mood directly rarely resu lt in effect sizes that reach or exceed the environmental context change effects. As an alternative, Glenberg (1997) suggest ed that information from the world is continuously represented unless suppressed, wh ich occurs in order to free up cognitive resources for more demanding tasks. In ot her words, environmental context cues are suppressed due to cognitive load. This idea informed two explanations of environmental context-dependent memory: the overs hadowing hypothesis and the outshining hypothesis. The overshadowing hypothesis (as describe d in Smith, 1988; Smith & Vela. 2001) states that context changes are due to a failure to encode environment contextual cues for a variety of reasons, such as control of attention or c onceptual processing of study materials. Like the term implies, if e nvironmental cues are overshadowed by other factors during encoding and are not stored, contextual change will not affect recall. Therefore, in the absence of factors that may overshadow environmental cues (e.g. material that allows for associative proce ssing or that draws attention away from the environment), a context change should be observed.

PAGE 19

9 Smiths (1978; 1979; 1985; 1988) outs hining hypothesis proposes that environmental cues are not used if other better cues are available. In other words, better non-contextual cues will outsh ine environmental cues much like the sun outshining all other celestial bodies in our sky. Unlike the overshadowing hypothesis, the outshining hypothesis states that environmen tal context cues are encoded, but are not u tilized during retrieval because better cues are available. Better cues can be defined as deeper processing of the material, generation of be tter retrieval cues due to deeper encoding strategies, or those that enc ourage inter-item associations. The overshadowing and the outshining hypothe ses may seem similar in that they both specifies failures of the environmental c ontext change effect, but they are actually different in their predictions of context cha nge. One distinction is that the overshadowing hypothesis is concerned with fa ilure to store contextual cu es at encoding and therefore there are none to use at retrieval. The out shining hypothesis suggests that contextual cues may be encoded successfully but the cues are not used at retrieval because better retrieval cues are available. Both are dependen t on associative processing and inter-item associations; however, overshadowing specifica lly explains context change effects as encoding effects while outshining specifically explains them as retrieval effects. These theories have been largely unsucce ssful at adequately predicting why and how context change effect occurs. Smith a nd Velas meta-analysis resulted in mixed findings for the overshadowing hypothesis and ou tshining hypothesis. It could be that these studies are not explor ing variables that may influence encoding and retrieval hence, affecting the magnitude of the context change effect. This may explain the mixed findings in the environmental context change literature. Therefore, I am proposing two

PAGE 20

10 variables that may influence the context change effect and should be empirically investigated: habituation and strategy Habituation How attention fluctuates over time is important in examining environmental context change. It has been shown that pa rticipants tend to mentally reinstate their environments by imagining the initial learning environment (Smith, 1979; 1984), resulting in a decrease in cont ext change effect. In the absence of such a strategy, context change effects are observed. Glenberg, Schroeder, and Robertson (1998) found that when participants gazes were diverted from the environment during retrieval, recall was better due to less processing devoted to th e present environment and more so to conceptual processing. Perhaps attentional cont rol is related to eff ectively processing and retrieving environmental cues. In Glenbergs (1997) theory of environmental suppression, a shifting of cognitive resources must occur for conceptual proces sing. These cognitive resources seem to be attentional resources. Ind eed, the overshadowing hypothe sis and outshining hypothesis both imply that if attention is drawn away from the learning or testing environment, no environment contextual manipulations should be observed. It would seem intuitive to state that most people are aware of their envi ronment, but that the attention fluctuates from the environment to the task, or focal in formation. The role of attention in context change studies seems to be under-explored and not examined thoroughly although most researchers agree that attention is important Experimental manipulations to make the environmental cues more prevalent (Godde n & Baddeley, 1975), novel or flashy (Dallet & Wilcox. 1968; Dulsky, 1935; Weiss & Ma rgolius, 1954), or included explicit instructions to note the e nvironment (Eich, 1985; Nixon & Kanak, 1981) have all been

PAGE 21

11 aimed to capture or direct attention. The environmental contexts for Godden and Baddeleys study were under water and on la nd. Participants in Dallet and Wilcoxs study were asked to wear a box of changing li ghts over their heads rather than change rooms. The box not only succeeded in producing a context change effect, but also caused nausea and disorientation for the participants. Most of us do not spend a large amount of our time in water, wearing scuba gear, or ad orning our heads with a box full of lights. The box of changing lights and the equipmen t for going under water (scuba diving gear) as well as aspects of the environment (physical suspension in water, physiological changes) were distinct and novel enough th at attention was more engaged to the environment than usual. These experimental manipulations contained unusual elements that drew attention. Environments oftentimes are familiar and frequented: we work at a certain place, have a home at a certain locati on, or visit preferred stores. On the other hand when an individual first encounters or enters a new environment, the environment is unfamiliar and novel; thus the individual attends or orients to it. After a length of exposure to this environment, the orienting response is re duced and habituation occurs (Cowan, 1988; Groves & Thompson, 1970; Sokolov, 1975). Dehab ituation or revival of the orienting response may appear if aspects of the envi ronment capture attention or attention are directed to the environment. The longer the participant spends in an environment, the less novel and interesting the e nvironment is. For example, waiting in the lobby at a new doctor's office, I swiftly surveyed the fu rniture arrangement and even find myself studying the prints hung against cheery wallpap er. As time went by (and it was obvious

PAGE 22

12 my doctor was running late), I was no longer in terested in the features of the room but instead perused the magazines available. Habituation occurs over time and predicts that the longer the time spent in an environment, the less likely environmental cues will be utilized. Therefore, when participants are asked to study information, th e environmental cues are less likely to be incidentally encoded with the study information. If the participan t immediately studies information upon entering the environment, th e participants atten tional resources may still be allocated to perceiving the envir onment. Therefore it would seem that participants who study information immediat ely would also incidentally encode environmental cues. When the environment at recall does not matc h the one at studying, these participants are more likely to show the phenomenon of forgetting due to mismatch of environments. Using the example above, if I was asked to remember a list of questions I wanted to ask my doctor during the time that I wa s still actively interested in the room, I would be more apt to forget the questions once I entered the doctor's examination room. Upon entering the lobby ag ain, I belatedly remembered the questions. If, however, I was asked to remember th e questions after I was no longer actively interested in my phys ical environment, I would be more likely to remember them upon entering the doctor's examination room. Le ss forgetting due to changing environments occurs because attention is no longer focused on the environment. Unfortunately, the literature on environmen tal context change has not extensively examined amount of time spent in an environm ent. Most real worl d events occur in a particular environment for more than the me re couple of minutes that are examined in laboratory studies. It would be more ecologi cally valid to examine the influence of time

PAGE 23

13 in an environment. A recent study (Isa rida, 2005) found that the longer time spent studying the words, the greater the magnitude of contextual information integrated with the study materials. Isarida proposed that en vironmental context effect is related to amount of time spent studying. This account is different from the habituation view in that Isarida predicts that time during encoding will increase the associative strength between the context and the item. The habitua tion view proposes that attention drawn to the environment before and during encoding results in sampling of the environment which will be incidentally encoded with the study materials. According to the overshadowing hypothesi s, the longer the time spent in the learning environment, the higher the chances of the contextual cues incidentally encoded with the studied materials. Smith and Velas meta-analysis did not show amount of time spent in an encoding environment influencing recall performance. Effect sizes were calculated and compared for studies that had examined environmental dependent memory. But these studies did not include amount of time spent in the environment prior to learning. Examined was time spent in the l earning environment either during learning, such as completing a distractor task in betw een presentation of words or after learning, such as completing a distractor task while still in the learni ng environment. No studies have yet looked at time spent in an environment prior to studying. Strategy Another variable that has not been extensive examined and may affect the environmental context change effect is part icipants initial encoding strategy. In their meta-analysis, Smith and Vela found that st udy materials that do not induce associative processing showed a greater context change effect whereas associative processing

PAGE 24

14 resulted in much smaller context depende ncy effects. These studies range from examining depth of processing (Murnane & Phelps, 1995; Smith, 1986; Smith, Vela, & Williamson, 1988) to varying the strength of retrieval cues of the study materials by associative processing. For example, in Smiths (1986) study, the shallow processing task was actually an incidental short term memory task in which participants heard a short list of words and after several sec onds, immediately recalled them or a deeper processing task in which participants were told to attempt to memorize the words. Afterwards, participants complete d a variety of distractor task s and then either recalled in the original or new environment. A contex t change effect was observed for the shallow processing but not for the deeper processing. Studies that have looked at inter-item associations a nd associative processing of study materials found that these reduced or eliminated the context change effect. Smith and Vela (2001) suggested that participants do no t utilize contextual cues even if they are available because associative processing pr ovides better retrieval cues: Recalling one item would guide the retrieval of another. This outcome can be explained by both the overshadowing and outshining hypothesis, but both attribute it to different reasons. The overshadowing hypothesis predicts that associative processing draws attention away from the environment and therefore environmental cues are not encoded and therefore are not available for use. The outshining hypothesi s predicts that environmental cues do get encoded but associative processing outshines th e environmental cues as retrieval cues and are, thus, utilized. Einstein and Hunt (1980) examined the levels of processing and organizational approach to encoding. They point out that levels of processing tasks often involve

PAGE 25

15 processing individual items whereas organizatio nal strategies involve processing between items. They conclude that there is a dis tinction between individua l item based processing and relational processing and that this dis tinction is important in regards to defining elaborative processing. Craik and Lockhart's (1972) concept of elaborative processing involves individual item processing where it ems are made more meaningful and hence more accessible in memory. According to th e levels of processing perspective, rote rehearsal is viewed as shallow processing because items are repeated in short term memory. Elaborate processing is rehearsing the items to store them into long term memory. The levels of processing view would argue that elaborative processing would result in better recall. However, Benjamin and Bjork (2000) found that in the presence of a time pressure, accessibility to items was mo re disrupted for elaborative rehearsal than rote rehearsal. They concluded that elabora tive rehearsal may facili tate stronger retrieval cues but there is a time trade-off. Ela borate rehearsal is effortful and resource consuming; engaging in elaborate rehearsal during encoding and resurrecting the mental framework during retrieval is time consuming (Masson & McDaniel, 1981). However, there is evidence suggesting th at the mere time and effort of elaborative processing of individual items is not enough to strengthen accessibility cues. Bradley and Glenberg (1983) found that time and attention spent dur ing rehearsing items i ndividually did not enhance recall. There was enhanced recall onl y when more than one item are rehearsed together, forming and strengtheni ng inter-item associations. Environmental context studies have only manipulated processing in the Lockhart and Craik sense. These studies utilizing a ssociative processing have found that this greatly reduces or eliminates context depe ndency. However, no studies have looked at

PAGE 26

16 encoding strategies that require relational item processing. In contrast to the Lockhart and Craik levels of processing view, depth of encoding strategies involve both forming inter-item associations as well as generating individual item distinctiveness. Therefore, shallow encoding would be viewed as main tenance rehearsal of individual items or rhyming the words. One form of deep enc oding involves generating an interactive image or a story for all the items (Bower, Clark, Lesgold, & Winzenz, 1969). This categorization of encoding strategies has been utilized by researchers examining recollection (Perfect & Das gupta, 1997), spacing effects (Delaney & Knowles, 2005), and directed forgetting (Sahakyan & De laney, 2003; Sahakyan, Delaney, & Kelley, 2004). Sahakyan and Delaney (2003) have found that changing internal contexts induces strategy changes in directed forgetting studies. In directed forgetting studies, participants are presented with two word lists and half are told to intentionally forget the first list. A typical outcome is the presence of both costs and benefits of directed forgetting, in which costs refer to the poorer recall that forget participants display for the first list relative to remember conditions and benefits re fer to the increased recall that forget participants display for the second list relative to the re member conditions. In a series of studies, Sahakyan and Kelley (2002) and Sahakyan (2004) have argued that the costs in directed forgetting can be attributed to a mental cont ext change. When instructed to forget the list, participants attempt to think of something else, which then changes their mental context. Therefore the context during enc oding of the first list and the context during recall are mismatched, leading to poorer recall fo r the first list or co sts. The benefits, however, are due to strategy changes that ar e induced by the mental context change.

PAGE 27

17 Sahakyan and colleagues (Sahakyan & Dela ney, 2003; Sahakyan, Delaney, & Kelley, 2004) have suggested that the me ntal context change that forget participants engage in are also evaluative in nature. Participants evaluate their encoding strategy for list one after being told to forget it and perhaps real ize that their current method is ineffective, hence a strategy change occurs for the second list. Without prior knowledge or training, participants tend to engage in shallow encoding, such as rote rehearsal (Delaney & Knowles, 2005; Sahakyan & Delaney, 2003; Sahakyan, Delaney, & Kelley, 2004). Therefore switching to a deeper encoding strategy would explain the increase in recall for the second list for forget part icipants or the observed benefits. They concluded that utilization of a deep encoding strategy, especially those that initially engaged in shallow encoding, can improve recall due to better encoding of the information and better retrieval cues. Deep encoding contains both relational and item-specific processing; generating an interactive image or story allows for each item to become linked or associated to another item. Retrieval of an item would th erefore guide recall of another item. Based on the above research findings, the type of strategy during encoding may account for context dependency. Deep encoding may pr ovide resistance to forgetting due to changing contexts. Because the majority of individuals without prio r training engage in shallow encoding, the context ch ange effect is observed. If instead deep encoding was used, the context change eff ect should be reduced or eliminated since associations between items serve as better retrieval cues than environmental cues. This predication would seem similar to the outshining hypothesi s in that there are better retrieval cues available and hence environmental cues ar e outshone, however the outshining hypothesis

PAGE 28

18 predicts that environmental cues are successf ully encoded. Engaging in an elaborative processing task may potentially interfere w ith encoding of environmental cues as cognitive resources are allocated to th e task, a suggestion of the overshadowing hypothesis. However, the overshadowing hypothesi s attributed the context change effect to this. Use of a shallow strategy is not as effortful and will not detract from encoding of environmental cues and use of a deep strate gy is more effortful and may detract from encoding of the cues. However, the lack of environmental cues is not necessary since the deep strategy provides better retrieval cues th an environmental context cues. It would be interesting to examine further these ideas as no studies in the environmental context literature, thus far, have extensively examined relational encoding strategies.

PAGE 29

CHAPTER 2 EXPERIMENT The purpose of the first experiment was to examine the effect of time spent in the encoding environment and recall of words in matched or mismatched environments while controlling for participants en coding strategies. Participants that utilize deeper processing such as inter-item associations usually display resist ance to contextual manipulations because recall of one item prim es the recall of another item (Smith, 1988; Smith & Vela, 2001; Steuck & Levys study as cited in Smith, 1988). In this sense, deep encoding of the material may result in items becoming more associated with one another and more meaningful in memory. When info rmation to be studied is more meaningful, environmental context effects have been observed to be lower than when the study materials are meaningful (Steuck & Levys stud y as cited in Smith, 1988; Velas study as cited in Smith, 1988). In light of this, experi mentally controlling participants encoding strategies would allow closer examination of the separate influences of time and participants encoding strategies. I exp ect to find evidence supporting the following predictions of the influence of habituation and deep encoding strategies on the magnitude of the environmental context change effect. Time and Strategy Predictions Overall, participants in the match conditions are expected to do better than those in the mismatch conditions, replicating the envi ronmental context change effect. This finding will replicate studies that found reliable context-change effect (see Smith & Vela, 2001) in that there is forget ting that occurs when changi ng to a physical environment 19

PAGE 30

20 different from that for encoding. Particip ants who study the words immediately in the experiment should have poorer recall than participants who waited seven minutes and then studied the words. This may be becau se participants who immediately studied the words may attenuate to the environment. In other words, those that study the words after a delay have habituated to the environment. Hence, they have more attentional resources to allocate to the task than t hose that have not habituated to the environment and are still attenuating. Participants trained through a deep encodi ng strategy are expected to have higher recall than those trained thr ough a shallow encoding strategy regardless of influence of time or if the environment matches. Partic ipants who were instru cted through a shallow encoding strategy should be affected by time spent in an environment prior to studying the words. In particular, recall should be worst for those who studied the words immediately than those who studied the wo rds after a delay when the environments mismatches, indicating a context change eff ect. For match conditions, there should be no influence of time since environmental cues are available for retrieval. There should also be a significant re call environment by times of encoding interaction. Participants that study words im mediately (early conditi on) in environments that match for encoding and recall should have the highest recall. Th e conditions that are expected to display the worst recall are th e ones in which participants study the words immediately and then change environments (late and mismatch environment condition). For participants who study the words after waiting (the late condition), those that are in environments that match for encoding and r ecall should do better than those that are in environments that mismatch. This findi ng will support what was earlier predicted:

PAGE 31

21 Habituation to the environment reduces the forgetting that occurs when environments mismatch. Participants who study words imme diately incidentally encode environmental cues with the target words, resulting in a dependence on environmental context cues. Therefore when the recall environment is diffe rent from the initial environment in which the words were studied, r ecall should be reduced. Working Memory Predictions Working memory was also included as a measure due to its role in attentional control (Kane & Engle, 2000). Delaney and Sahakyan (2004) demonstrated that mental context change is related to working memory capacity. They instructed participants to continue to remember a list of words a nd then a mental context change task was administered. They found that high working memory participants were more affected by a mental context change, forgetting more info rmation than low span participants. Based on these findings, I expect individual differenc es in working memory span to be related to recall in conditions where there is a change of environment but not in conditions where there is no environment change. In particul ar, high span working memory individuals recall may be reduced when there is a physic al environmental change than low span individuals. If this finding was supporte d, consistent with Delaney and Sahakyans (2004) results, high span working memory i ndividuals would, thus, be more contextdependent than low span individuals. Working memory is expected to signif icantly correlate with recall from experimental conditions (i.e., strategy, time a nd environment manipulations). Those with higher working memory spans will be more likely to spontaneously use deep encoding strategies. Encoding strategi es are effortful and maintain ing these processes after an environment has changed may require the part icipant to inhibit attenuating to the new

PAGE 32

22 environment and resurrect the mental scaffo ld of the effortful encoding strategy during recall. Individuals with high working memo ry span should have decreased performance when environments are changed. However individuals with relatively low working memory span should remain unaffected by the environment changes. This may be because those with low working memory span may not need to inhibit the environment while engaged in effortful encoding strategies. Time is not expected to have a significant correlation with working memory. There is no indication from the research literature that suggests that habituation is influenced by wo rking memory span. Therefore, the effects of the length of time spent in the encoding environment should probably be unaffected by working memory span. Methods In this experiment, the following variab les were manipulated: initial encoding strategies, duration of time spent in an e nvironment prior to learning, and whether the environment for learning and recall matched or mismatched. It has been argued that salient elements of the initial environment may still be pres ent in the second environment, drastically reducing the environmental cont ext change effect (Fernandez & Glenberg, 1985; Smith & Vela, 2001). For instance, the laboratory room for th is environment is very similar to a lobby or office; both include similar office furniture and arrangements or similar presentation instruments. Therefore, to maximize contextual change, care is taken to ensure no part of the in itial environment is salient wh en changing environments. In this experimental proposal, environmental context change is manipulated when the environment is changed from an outside to inside location and vice versa, different

PAGE 33

23 experimenters for each phase of the change, and different methods of studying the words (on a multimedia player on the computer versus on a portable stereo). Delaney and Knowles (2005) found that for participants engaged in shallow encoding, the most frequent strategy reported was verbal rehearsal and for deep encoding was a story generation mnemonic. In De laney and Knowless Experiment 2, they experimentally controlled participants en coding strategies by training participants through a shallow or deep encoding strategy and requiring the usage of this strategy throughout the experiment. A similar method was used to fix participants initial encoding strategy in this experiment. Participants Participants were University of Florid a undergraduates who received course credit or extra credit for completing the experiment. Participants were tested individually with 8 participants in each of the eight experiment al conditions for a total of 64 participants. Materials Twenty unrelated, medium-frequency English nouns were selected. The words were recorded by a male voice at a rate of 1 word per 4 s and edited on Adobe Audition 1.5. The inside environment is a computer lab room inside the psychology building with a computer, two chairs, file cabinets and shelves, various computer equipment on the ground and shelves, books and stacks of pape r on the shelves, and white walls. The outside environment is a relatively seclude d and quiet picnic place outside the psychology building. This picnic place is su rrounded by trees, has a roof over it and includes a stone table and stone benches on two sides of the table. The words were presented on a multimedia player on the computer for the inside condition. For the outside condition, the words were presented on a portable ster eo with a compact disc

PAGE 34

24 player. For the purposes of this experiment different experimenters were considered different for only one variable: gender. Gender was considered a substantial enough difference for conditions where the environmen t was changed. Therefore, the gender of experimenters was counterbalanced to control for possible biases. Also, care was taken to ensure that the experimenters appearance were not similar (e.g. clothes, hairstyle). Procedure Encoding Participants were tested i ndividually. Before they began the experiment, informed consent was obtained from the participant in th e hallway or neutral zone. Then they were led to the encoding environment and asked to sit facing away from the experimenter to discourage focusing solely on the experimenter. Half of the participants were told the experimenter was still se tting up the experiment (the late condition). They were told to just sit quietly and wait until the experimenter finished preparing the experiment. The experimenter was actually surfing the internet on the computer for the inside condition or engaged in scoring data or a search-a-word game for the outside condition to appear occupied. The participant was discoura ged from studying or from engaging in conversation with the experimenter to give the participant an opport unity to observe the environment. After seven minutes, the particip ant was told that the experiment was ready to begin and was then trained through a shal low or deep encoding strategy. The other half of the participants, upon entering the encoding environment, was immediately asked if they were ready to begin the experiment and trained through either a shallow or deep encoding strategy. Participants in the shallow condition were instru cted in the verbal rehearsal encoding strategy prior to studying the words. They we re instructed to rehearse the words out loud, adding each new word to th e set already rehearsed. They were also

PAGE 35

25 told not to be alarmed if they forgot any wo rds and to continue rehearsing the words that they do remember. Participants in the deep encoding condition were instructed in the story generation encoding strategy where they were required to generate aloud a story utilizing each word. They were also told that they are required to use all the words in the story. After studying the words, participants e ngaged in a math distractor task for 90 s to reduce recency effects. Recall Participants in the match environment condition were led to the neutral zone and then were led back to the original environment by the same experimenter. For the mismatch environment condition, participants were led to the outside or the inside environment depending on their initial encoding environment and a different experimenter conducted the next portion of the experiment. Time between changing environments was controlled for both match a nd mismatch conditions so that the time for walking the participant to the new environmen t and the time for walking the participant to a neutral zone and back to the same environment is the same: 90 seconds. They were then asked to free recall the words on a bl ank sheet of paper for 90 seconds. Post-recall measures Retrospective verbal reports were collected to determine participants initial affective states and whether it changed and what they were thinking for those in the late condition. They were also asked if the strategy task was difficult (i.e. following instructions or engaging in th e task) and if the words were spoken clearly and audibly. After this, all participants completed a working memory span task called the Triangle C Span Task. The Triangle C Span Task invol ves viewing slides on Microsoft PowerPoint that consist of light purple tria ngles, dark purple triangles, an d light purple squares. Their

PAGE 36

26 task is to count the light purpl e triangles out loud and rememb er this total. They will view 2 6 slides before being asked to reca ll the sequence of total number of light purple triangles on each slide. Triangle C Span scor es range from 0 58. Figure 1-1 illustrates example slides. (a) (b) (c) Figure 1-1. Triangle C Span Task to Asse ss Working Memory Span Example Slides. Participants should count and remember 4 pur ple triangles in the (a) first slide and 8 purple triangles in the (b) s econd slide. When the prompt to recall the total number of purple triangles from each slide appear s (c), they should recall 4 and 8. Results Time, Strategy, and Environment Variables In order to ensure proper counterbala ncing, variables such as gender of experimenters and environment order were included in all main analyses; but since there are no significant interactions, were collapsed over. Mean proportions of words were calculated. A 2 Time of Encoding 2 En coding Strategy x 2 E nvironment betweensubjects ANOVA with match and mismatch envi ronments for the first factor, early and later encoding times as the second factor, and encoding strategy as the third factor revealed no significant main effects for time F (1, 56) =.008, MSE = .0002, p = .929,

PAGE 37

27 environment F (1, 56) = 2.031, MSE =.040, p = .160, or significant interactions: for time and strategy, F (1, 56) = .0006, MSE = .032, p = .859; for time and environment, F (1, 56) = 2.570, MSE =.0510, p = .115; for strategy and environment, F (1, 56) = 1.785, MSE = .0351, p = .187; for time and strategy and environment, F (1, 56) = .198, MSE = .0039, p = .658. The main effect of strategy was significant, F (1, 56) = 28.555, MSE =.563, p < .001. Participants in the deep encoding stra tegy have higher recall than those in the shallow encoding strategy (see Table 1-1 for overall means and standard deviations) regardless of time or if the environment matches or mismatches. Table 1-1. Recall Mean Proportions and Standa rd Deviations for Early, Late, Match, Mismatch, Shallow Encoding, and Deep Encoding Conditions Strategy Shallow Deep Time Mean S.D. Mean S.D. Early Match .306 .073 .519 .189 Mismatch .343 .018 .494 .201 Late Match .344 .132 .600 .136 Mismatch .300 .046 .431 .194 Working Memory Nine participants did not have working me mory span scores because their scores were excluded due to computer failures or failure to follow instructions. These nine participants were replaced. Participants Tr iangle C Span scores ranged from 9 to 45,

PAGE 38

28 with a mean of 25.42 ( SD = 9.307). There were no signifi cant correlations between working memory and recall of the words, r = .002, p = .988, which was not what was predicted. To investigate the possibility that recall in the experimental conditions was dependent on individual differences in working memory, linear regression analyses were employed (Jaccard and Turrisi, 2003). Linear regression has been utilized by Delaney and Sahakyan (2004) to examine individual di fferences in working memory for mental context change effect. Linear regression an alyses were conducted to examine proportion of words recalled as a function of working memory, experimental conditions (time, environments), and their interaction. Expe rimental condition was entered as orthogonal contrast-coded variables and the analyses were conducted separately for each strategy group ( shallow and deep ). The main effects were entered simultaneously and the interaction terms were entered to determin e if any of the interactions explain any additional variance. The total model was significant for shallow encoding, F (6, 25) = 2.940, p < .05 but not for deep encoding, F (6, 25) = .794, p = .583. For shallow encoding, there was a significant main effect of environment, F (1, 25) = 2.805, p < .01 and revealed a significant work ing memory and time interaction, F (1, 25) = 2.340, p < .05. Thus, the effect of experimental condi tions was modulated by working memory span when participants were required to engage in a shallow encoding strategy. For shallow encoders, the higher the working memory, the higher the proportion of words recalled for matching; whereas for mismatch environmen ts, differences in working memory span does not influence recall envir onments (see Figure 1-2).

PAGE 39

29 10 20 30 40 50WM 0.20 0.30 0.40 0.50 0.60Recall Condition match mismatch Strategy: shallowR Sq Linear = 0.03 R Sq Linear = 0.316 Figure 1-2. Best-fitting Regre ssion Lines for Working Memory and Proportion of Words Recalled in Each Environment for Shallow Encoding Condition For deep encoders, working memory is not influenced by physical environmental manipulations (see Figure 1-3).

PAGE 40

30 10 20 30 40WM 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90Recall Condition match mismatch Strategy: deepR Sq Linear = 0.076 R Sq Linear = 0.022 Figure 1-3. Best-fitting Regre ssion Lines for Working Memory and Proportion of Words Recalled in Each Environmen t for Deep Encoding Condition Retrospective Reports For the retrospective reports, participants were asked a bout their affective states before the experiment, after the delay (for those in the late condition), and what they were thinking about during the delay. Thes e are experimental checks that ensure that participants recall is not bette r or poorer due to high arousal Past research on emotion and arousal has shown that higher arousal tends to result in better memory performance (Anderson, 1988; Duffy, 1962; Mandler, 1975 ; Yerkes & Dodson, 1908). In this

PAGE 41

31 experiment, participants responses were categorized as either high arousal (e.g. excited, impatient, tense, nervous, anxious), neutra l (e.g. feels fine, good, curious, relaxed, calm, normal), or low arousal (e.g. sad, tired, sleepy, bored). When the experiment started, 35.9% participan ts experienced high arousal, 48.4% experienced ne utral arousal, and 12.5% experienced low arousal. The affec tive states did not significantly affect recall; in other words, despite the past findings on arousal and better memory performance, participants performed at a comparable rate whether they were highly aroused, neutral, or had reduced arou sal. Of those that were in the late condition, 53.1% had different affective states before and af ter the delay (52.94% changed to a high arousal state, 23.53% changed to a neutral arousal state, and 23.53% change d to a low arousal state). However, this switching did not signifi cantly affect recall. Although participants found engaging in the strategy task was slightly difficult, this was more due to the length of the word list when adding the words to the cycle of repetitions rather than any difficulties with following the instructions No participants reported having any difficulties hearing the words clearly. Post Hoc Analyses Of the 64 participants that partic ipated in the experiment, 48.4% (n = 31) were from introductory psychology courses. The others, 51.6% ( n = 33), were from upper division psychology courses. This may have biased the results; to test this post hoc, descriptive statistics for the data was anal yzed separately and the mean proportions of recall were compared for those in the intr oductory psychology course s and those in upper division psychology courses. There were 2-6 of each type of student in each condition cell; therefore, because the cells do not contai n equal sample numbers, the data should be interpreted with caution. However, the upper division psychology students had overall

PAGE 42

32 slightly higher recall for both shallow and deep encoding than introductory psychology students. This is especially true for deep encoding for some of the conditions, such as early and match M = .388 (SD = .085) for introductory students versus M = .650 ( SD = .173) for upper division students (s ee Table 1-2 and Table 1-3). Table1-2. Recall Mean Proportions and Standard Deviations for Introductory Psychology Course Students Strategy Shallow Deep Time Mean S.D. Mean S.D. Early Match .250 .050 .388 .085 Mismatch .342 .020 .450 .354 Late Match .276 .035 .608 .139 Mismatch .300 .000 .438 .144

PAGE 43

33 Table1-3. Recall Mean Proportions and St andard Deviations for Upper-Division Psychology Course Students Strategy Shallow Deep Time Mean S.D. Mean S.D. Early Match .340 .065 .650 .173 Mismatch .350 .000 .508 .174 Late Match .367 .147 .625 .160 Mismatch .300 .071 .475 .203

PAGE 44

CHAPTER 3 GENERAL DISCUSSION Environment Past research has shown that context cha nge effect is a modest, although reliable effect (Smith & Vela, 2001). However, no significant effect of environment was observed; participants did equally well if the encoding and retr ieval environments matched or mismatched. Ferdnandez and Gl enberg (1985) did not find any reliable context change effects and actually observed a different context advantage as frequently as not finding any advantage. They suggested that differences in environments and type of disruption tasks in the literature may re sult in discrepancies in the magnitude of context change effect. Perhap s the context change effect does occur but is undetectable by current experimental manipulations. This explanation, however, seems unlikely. Other researchers have been able to find context change effects (Smith, 1979; 1984; 1986; Smith, Glenberg, & Bjork, 1978). The environmental context change effect is also vulnerable to a number of different manipulations, as was discussed in Smith a nd Velas meta-analysis. In this study, participants in the match condition were asked to walk to a neutral zone and then back to the initial environment. In several of Smiths studies (1979, 1984, 1986), the participants in the match condition were asked to wait in a neutral zone. This difference in disruption task may account for no significant context chan ge effects in the present study: Smiths participants, while waiting in the neutral z one, may view this as an opportunity to rehearse the studied words. Whereas part icipants in this study were continuously 34

PAGE 45

35 walking for the duration of the disruption ta sk and may view this task as beginning another unrelated task. Therefore these participants are less likely to engage in rehearsal. A criticism of past studies that did not find environmental context change effects was that the environments we re not different enough. Care was taken in this study to maximize the differences between the two environments by choosing an indoor office setting and an outdoor picnic table se tting and the experimenters conducting the environments. However, participants could be viewing the environmen ts as part of the same task (studying words) and therefore, thei r expectations did not alter. Manipulating participants expectations is a difficult experimental control. Isarida (2005) found that type of task can contribute to context cha nge effect, yet environm ental context change effect has been demonstrated for same-task studies (see introduction). Encoding Strategy There was a significant effect of strate gy on recall. Participants that were instructed to engage in a deep encoding strategy had higher recall th an those that were instructed through a shallow en coding strategy. Past environmental context research found that deep processing resulted in be tter recall (Murnane & Phelps, 1995; Smith, 1986; Smith, Vela, & Williamson, 1988) possibly due to relational inter-item processing. Therefore, an associative inter-item processi ng that converts the targ et items into more meaningful information will result in better recall, as was observed in this study. This replicates past research findings where deep encoding strategies result in better recall than shallow encoding strategies (Delan ey & Knowles, 2005; Sahakyan & Delaney, 2003; Sahakyan, Delaney, & Kelley, 2004). It ma y be that initial encoding strategy may be powerful enough to eliminate any environmen tal or time influences. If this was the

PAGE 46

36 case, then it would be interesting to ex amine shallow encoding strategy for future research as the majority of partic ipants engage in shallow encoding. Time It would seem from real-life observati ons that the more time spent in an environment, the easier and better developed the representation of the environment will be. Smiths overshadowing hypothesis (as de scribed in Smith, 1994; Smith and Vela, 2001) actually predicts that th e longer the time spent in an environment, the higher the chance that environmental retrieval cues will be utilized. Therefore, changing environments will result in poorer recall because of the util ization and greater dependency of environmental cues. Regardle ss, time spent in an environment, whether enhancing or reducing opportunities to utilize environmental context cues, seems intuitively to significantly influence the magnitude of context change effect. Surprisingly, there were no significant time manipulation effects; there were no support for the habituation hypothesis as an adequate explanation of the c ontext change effect. One possible explanation may be that the time interval in this study (seven minutes) was not enough for participants to habituate to the environment. Pe rhaps participants did not attenuate to the environments as was hope d and had zoned out during the delay. It does not seem to be that the environments must be radical to catch pa rticipants attention; however, it would seem that pa rticipants are more focused on completing the experiment and therefore their attenti on becomes tunnel-vision. Working Memory There was mix support for the working memory predictions. There were no significant correlations for working memory span and word recall, suggesting that perhaps the environmental context change effect is not related to working memory. This

PAGE 47

37 does not provide direct support for the predicti on that individual di fferences in working memory should also reflect and predict the magnitude of environmental context change effect, specifically, that high span working me mory individuals should be more affected by changing environments. Delaney and Sahakyan (in press) found that when participants were asked to forget informa tion, high span individuals had poorer recall compared to low span individuals. They attrib uted this to the mental context change that occurs with a forget cue; high span indi viduals were more affected by the mental context change than low span individuals. Surprisingly, this was not observed with regular correlations as was predicted. However, multiple linear regression analyses revealed interesting trends. Based on the results of the linear regression analyses to assess individual differences in working memory affecting recall of experimental c onditions, it would seem that there was some support for the hypotheses. For those engaged in a shallow encoding strategy, those with high working memory span were more affected by the manipulations of environment and time, resulting in lower recall when environm ents matched than mismatched, replicating Delaney and Sahakyans (2004) findings. For those engaged in a deep encoding strategy, their individual differences in working memory did not influence experimental manipulations, and hence did not affect recall. These findings provide some support for the prediction that environmental and time manipulations should affect those in the shallow encoding condition but not for those in the deep encoding condition. Engle and Turner (1986, 1989) argued that working memory is independent of task-related processing. However, it woul d seem more likely that deep encoding provides such powerful retrieval cues that there is no influenc e of individual differences

PAGE 48

38 in working memory span. Kane, et al (2000, 2001) argued that working memory capacity was attentional based. Due to the mixed working memory results, this also suggests that perhaps environmental cont ext change effect is less dependent on attentional processing and is dr iven much more by strategy than was previously thought. However, the trends are in the predicted direction. Population Sample Problems The retrospective reports s howed that participants te nd to become increasingly more aroused after a delay (e.g. impatient, te nse, nervous, excited). However, most participants initially begin the experiment as either aroused or neutral (e.g. normal, good, fine, curious). Interestingly, participants affective states did not affect recall, which does not support Eichs (199 5) mood hypothesis theory. The non-significance of the results may be due to a number of these reasons. However, it would seem that there were pr oblems with the sample population. Upper division psychology students may have more expert knowledge and hence, may have guessed the purpose of the study. This ma y actually affect the results of the late condition in which participants were told to wa it quietly while the e xperimenter is setting up; students from upper division psychology co urses may see through the ruse. These students from upper division c ourses may also be better at engaging in the encoding strategies since they may have already developed better study stra tegies. The post hoc comparisons suggest that there were differe nces in the performance of introductory students versus upper division students. Ther e were differential performances for the participants depending on whether they we re introductory psychology or upper division students as is evidenced by very different m eans. However, due to the low and unequal

PAGE 49

39 number of participants in each cell, caution should be ta ken when interpreting these results. It could also be possible that there were out liers in the data set, as is evidenced by high standard deviations in some conditions such as the early, mismatch and deep experimental conditions for introductory to psychology students (SD = 7.071). These outliers may have biased the data. In future follow-up studies, these outliers should be replaced. Another potential explanation for the non-si gnificant results is not having enough statistical power to detect the effects. Th ere were eight particip ants in each of the experimental cells which may not be enough participants to det ect context change effects. In spite of this, there were no ceiling effects or flooring effect s. Follow-up studies should include more participants and be cau tious of participant recruitment. General Conclusion Despite all these possibilities, it could be that initial encoding strategy greatly influences context change effect It would make sense to th ink that information that is more meaningful is less forgotten. In the real -life evidence, we often find that we forget what we initially set out to do when moving to a new environment, such is the case when going to obtaining a drink while watching televi sion or retrieving an office supply in the midst of typing. However, perhaps more of ten, we do not forget that we are going to work or class as soon as we leave our homes or that we are going to return an important phone call in the midst of the day. Perhaps this is because this information is meaningful; we usually have a routine where we go to work or class at the same ti me every day or that a phone call is very important. This inform ation would be deeply encoded because it is organized, important, and meaningful. The items or task that is usua lly forgotten due to

PAGE 50

40 an environmental context change effect may be not as important as other activities that are occurring; getting more paper for the printer or get ting a drink in a midst of a broadcast game may be not as important or of ones interest as working on the paper that is due soon or the game that is playing on the television. We unconsciously prioritize our tasks and information and hence, the items that are not as important are not processed as much to prevent exhausting cognitive resources. It is interesting to note th at encoding strategies play a bigger role in incidental memory, such as environmental context change than was originally expected. In this study, participants encoding strategies were controlled; participants were required to utilize a deep or shallow encoding strategy to study the words. It is important to remember that without instru ction, participants tend to overwhelmingly choose an initial shallow encoding strategy (Delaney & Knowles, 2005; Sahakyan & Delaney, 2003; Sahakyan, Delaney, & Kelley, 2004). This may not be adequate to explain the discrepancies in the environmental context chan ge literature, but future research should examine the role of encoding strategies a nd the magnitude of environmental context change effect.

PAGE 51

LIST OF REFERENCES Abernethy, E. M. (1940). The effects of changed environmental conditions upon the results of college examinations. Journal of Psychology, 10, 293-301. Anderson, C. A. (1976). Coping behaviors as in tervening mechanisms in the inverted-U stress-performance relationship. Journal of Applied Psychology, 61, 30-34. Anderson, J. R., Bjork, R. A., & Bjork, E. L. (1994). Remembering can cause forgetting: Retrieval dynamics in long-term memory. Journal of Experimental Psychology: Learning, Memory, & Cognition, 20, 1063-1087. Baddeley, A. D. (1982). Do mains of recollection. Psychological Review, 89, 708-729. Benjamin, A. S., & Bjork, R. A. (2000). On the relationship between recognition speed and accuracy for words rehearsed via rote versus elaborative rehearsal. Journal of Experimental Psychology: Learning, Memory, & Cognition, 26, 638-648. Bjork, R. A., & Richardson-Klavehn, A. ( 1988). On the puzzling relationship between environmental context and human memory. In C. Izawa (Ed.) Current issues in cognitive processes: The Tulane Flowerree Symposium on Cognition (pp. 313344). Hillsdale, NJ: Lawrence Erlbaum. Bower, G. H., Clark, M. C., Lesgold, A. M., & Winzenz, D. (1969). Hierarchical retrieval schemes in recall of categorized word lists. Journal of Verbal Learning and Verbal Behavior, 8, 323-343. Bradley, M. M., & Glenberg, A. M. (1983) Strengthening associations: Duration, attention, or relations? Journal of Verbal Learning & Verbal Behavior, 22, 650666. Cowan, N. (1988). Evolving conceptions of memo ry storage, selectiv e attention, and their mutual constraints with the human information-processing system. Psychological Bulletin, 104, 163-191. 41

PAGE 52

42 Craik, F. I., & Lockhart, R. S. (1972). Levels of processi ng: A framework for memory research. Journal of Verbal Learning & Verbal Behavior, 11, 671-684. Craik, F. I., & Tulving, E. ( 1975). Depth of processing and the retention of words in episodic memory. Journal of Experimental Psychology: General, 104, 268-294. Dallett, K., & Wilcox, S. G. (1968). Contextual stimuli and proactive inhibition. Journal of Experimental Psychology, 78, 475-480. Delaney, P. F., & Knowles, M. E. (2005). En coding strategy changes and spacing effects in the free recall of unmixed lists. Journal of Memory & Language, 52, 120-130. Delaney, P. F., & Sahakyan, L. (2004). Unexpected costs of working memory: The impact of contextual change. Manuscript submitted for publication. Duffy, E. (1962). Activation and behavior New York: Wiley. Dulsky, S. G. (1935). The effect of a cha nge of background on recall and relearning. Journal of Experime ntal Psychology, 18, 725-740. Earles, J. L., Smith, A. D., & Parks, D. C. (1994). Age differences in the effects of facilitating and distract ing context on recall. Aging & Cognition, 1, 141-151. Eich, J. E. (1980). The cue-dependent nature of state-dependent retrieval. Memory & Cognition, 8, 157-173. Eich, (J.) E. (1985). Context, memory and integrated item/context imagery. Journal of Experimental Psychology: Learning, Memory, & Cognition, 11, 764-770. Eich, (J.) E. (1995). Mood as a me diator of place-dependent memory. Journal of Experimental Psychology: General, 124, 293-308. Eich, J. E., Weingartner, H., & Stillman, R. C. (1975). State-dependent accessibility of retrieval cues in the reten tion of a categorized list. Journal of Verbal Learning & Verbal Behavior, 14, 408-417. Einstein, G. O., & Hunt, R. R. (1980). Leve ls of processing and organization: Additive effects of individual-item and relational processing. Journal of Experimental Psychology: Human Learning & Memory, 6, 588-598. Fernandez, A., & Glenberg, A. M. (1985) Changing environmental contexts do not reliably affect memory. Memory & Cognition, 13, 333-345. Glenberg, A. M. (1997). What memory is for. Behavioral & Brain Sciences, 20, 1-55.

PAGE 53

43 Glenberg, A. M., Schroeder, J. L., & Robertson, D. A. (1998). Averting the gaze disengages the environment and facilitates remembering. Memory & Cognition, 26, 651-658. Godden, D. R., & Baddeley, A. D. (1975). C ontext-dependent memory in two natural environments: Land and underwater. British Journal of Psychology, 66, 325-331. Godden, D. R., & Baddeley, A. D. (1980). When does context influence recognition memory? British Journal of Psychology, 71, 99-104. Groves, P. M., & Thompson, R. F. (1970) Habituation: A dual-process theory. Psychological Review, 77, 419-450. Isarida, T. (2005). Study time effect on free recall within and out of context. Memory, 13, 785-795. Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression. Thousand Oaks, CA: Sage. Kane, M. J., Bleckley, M. K., Conway, A. R. A., & Engle, R. W. (2001) A controlledattention view of working memory capacity. Journal of Experimental Psychology: General, 130, 169-183. Kane, M. J., & Engle, R. W. (2000). Worki ng-memory capacity, proactive interference, and divided attention: Limits on long-term memory retrieval. Journal of Experimental Psychology: Learning, Memory, & Cognition, 26, 336-358. Krafka, C., & Penrod, S. (1985). Reinstatemen t of context in a field experiment on eyewitness identification. Journal of Personality & Social Psychology, 49, 58-69. Mandler, G. (1975). Mind and emotion New York: Wiley. Masson, M. E., & McDaniel, M. A. (1981). The role of organizational processes in longterm retention. Journal of Experimental Ps ychology: Human Learning and Memory, 7, 100-110. Murnane, P., & Phelps, M. P. (1995). Effects of changes in relative cue size strength on context-dependent recognition. Journal of Experimental Psychology: Learning, Memory, & Cognition, 21, 158-172. Nixon, S., & Kanak, N. J. (1981). The intera ctive effects of instructional set and environmental context changes on the serial position effect. Bulletin of the Psychonomic Society, 23, 139-142.

PAGE 54

44 Park, D. C., Smith, A. D., Morrell, R. W., Pug lisi, J. T., & Dudley, W. N. (1990). Effects of contextual integration on reca ll of pictures in older adults. Journal of Gerontology, 45, 52-57. Perfect, T. J., & Dasgupta, Z. R. R. ( 1997). What underlies the deficit in reported recollective experience in old age?. Memory & Cognition, 25, 849-858. Sahakyan, L. (2004). The destructive eff ects of the forget instructions. Psychonomic Bulletin & Review, 11, 555-559. Sahakyan, L., & Delaney, P. F. (2003). Can en coding differences explain the benefits of directed forgetting in the list-method paradigm?. Journal of Memory & Language, 48, 195-201. Sahakyan, L., Delaney, P. F., & Kelley, C. M. (2004). Self-evaluation as a moderating factor of strategy change in directed forgetting benefits. Psychonomic Bulletin & Review, 11, 131-136. Sahakyan, L., & Kelley, C. M. (2002). A contextual change account of the directed forgetting effect. Journal of Experimental Ps ychology: Learning, Memory, & Cognition, 28, 1064-1072. Saufley, W. H., Jr., Otaka, S. R., & Bavaresc o, J. L. (1985). Contex t effects: Classroom tests and contextual independence. Memory & Cognition, 13, 522-528. Smith, A. D., Park, D. C., Earles, J. L. K., Shaw, R .J., & Whiting, W. L. (1998). Age differences in context integration in memory. Psychology & Aging, 1, 21-28. Smith, S. M. (1979). Remembering in and out of context. Journal of Experimental Psychology: Human Learning & Memory, 5, 460-471. Smith, S. M. (1982). Enhancement of recall us ing multiple environmental contexts during learning. Memory & Cognition, 10, 405-412. Smith, S. M. (1984). A comparison of two techniques for reducing context-dependent forgetting. Memory & Cognition, 12, 477-482. Smith, S. M. (1985a). Effects of number of study environments and learning instructions on free-recall clustering and accuracy. Bulletin of Psychonomic Society, 23, 440442. Smith, S. M. (1985b). Environmental contex t memory and recognition reconsidered. Bulletin of Psychonomic Society, 23, 173-176. Smith, S. M. (1986). Environmental context-dependent recognition memory using a short-term memory task for input. Memory & Cognition, 14, 347-354.

PAGE 55

45 Smith, S. M.(1988). Environmental context-depe ndent memory. In G. M. Davies & D.M. Thomson (Eds.), Memory in context: Context in memory (pp. 13-34). Oxford, England: John Wiley & Sons. Smith, S. M. (1995). Getting into and out of mental ruts: A theory of fixation, incubation, and insight. In R.J. Sternb erg & J.E. Davidson (Eds.), The nature of insight (pp. 229-251). Cambridge, MA, US: The MIT Press. Smith, S. M., Glenberg, A., & Bjork, R. A. (1978). Environmental context and 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. Smith, S. M., Vela, E., & Williamson, J. E. (1988). Shallow input processing does not induce environmental context-dependent recognition. Bulletin of Psychonomic Society, 26, 537-540. Sokolov, E. N. (1975). The neuronal mech anisms of the orienting reflex. In E.N. Sokolov & O.S. Vinogradova (Eds.), Neuronal mechanisms of the orienting reflex (pp. 217-235). Hillsdale, NJ: Lawrence Erlbaum. Strand, B. Z. (1970). Change of context and retroactive inhibition. Journal of Verbal Learning & Verbal Behavior, 9, 202-206. Tulving, E., & Thomson, D. M. (1973). Encodi ng specificity and retrieval processes in episodic memory. Psychological Review, 80, 359-380. Turner, M. L., & Engle, R. W. (1986). Working memory. Proceedings of the Human Factors Society, 30, 1273-1277. Turner, M. L., & Engle, R. W. (1989). Is working-memory capacity task dependent? Journal of Memory and Language, 28, 127-154. Watkins, M. J. (1978). Engrams as cuegrams and forgetting as cue-overload: A cueing approach to the structure of memory. In C.R. Puff (Ed.) The structure of memory (pp. 347-372). New York: Academic Press. Watkins, O. C., & Watkins, M. J. (1975). Build-up of proactiv e inhibition as a cueoverload effect. Journal of Experimental Psychology: Human Learning & Memory, 1, 442-452. Weingartner, H., Adefris, W., & Eich, J. E. (1976). Journal of Experimental Psychology: Human Learning & Memory, 2, 83-87.

PAGE 56

46 Weiss, W., & Margolius, G. (1954). The effect of context stimuli on learning and retention. Journal of Experimental Psychology, 48, 318-322. Wickens, D. D. (1970). Encoding categories of words: An empirical approach to meaning. Psychological Review, 77, 1-15. Yerkes, R. M., & Dodson, J. D. (1908). The relati on of strength of stimulus to rapidity of habit-formation. Journal of Comparative Neurology and Psychology, 18, 459482.

PAGE 57

BIOGRAPHICAL SKETCH I was born in Thailand in 1982; at that ti me, my parents and br other were on their way to America. They were the boat peopl e of Vietnam, refugees of the Vietnam War in search of a better home and a hopeful future. Three months later, we eventually ended up in Oklahoma City, Oklahoma, which I would come to regard as my home city and would graduate from Northwest Classen High School in 2000. In 2003, I graduated from Oklahoma State University with a Bachelor of Science degr ee in psychology and a minor in biology. I entered the cognitive psychology program at the University of Florida in 2003. 47


Permanent Link: http://ufdc.ufl.edu/UFE0015866/00001

Material Information

Title: Influence of Time Spent in an Environment and Encoding Strategies on the Environmental Context Change Effect
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015866:00001

Permanent Link: http://ufdc.ufl.edu/UFE0015866/00001

Material Information

Title: Influence of Time Spent in an Environment and Encoding Strategies on the Environmental Context Change Effect
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015866:00001


This item has the following downloads:


Full Text
















INFLUENCE OF TIME SPENT
IN AN ENVIRONMENT AND ENCODING STRATEGIES
ON THE ENVIRONMENTAL CONTEXT CHANGE EFFECT










By

KHANH N. NGHIEM


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Khanh N. Nghiem















ACKNOWLEDGMENTS


It has been a long journey and I would like to thank all those who have helped me

along the way. I thank my advisor, Dr. Peter Delaney, for his guidance and providing me

with the necessary motivation to see this to the end. I would also like to thank my

committee members Dr. Ira Fischler and Dr. Scott Miller, for their valuable input,

support, and patience throughout the duration of this thesis. I am deeply indebted to

Lauren Saunders for her hard work and dedication in managing the lab and assisting me.

I gratefully thank Jason Bendezu and Manuel Lopez for helping me collect a much

needed portion of the data on very short notice. Michael Kung helped me with the study

materials and I thank him for his assistance. I also thank the following people for helping

me collect the pilot data and experimental data: Chris Cardani, Jaye Murray, Matt Smith,

Stevie Fisher, and Alfredo Lloreda. I would like to thank Keith McGregor for being an

unshakeable supporter. Dr. Carolyn Tucker has been a professional inspiration and I

would like to thank her for her advice, understanding, and patience when I needed time

away from work to collect data, write, and think.

Last but not least, I would like to thank the people in my personal life who may not

have directly contributed to the completion of this thesis, but nonetheless are important to

me. I would like to thank my parents who have supported my decisions and have been

inspirational; I appreciate the sacrifices they have made in order to provide my brother

and me a better future and education. I would like to thank my brother, who is my best















friend and staunchest supporter. I am grateful for my best friend, Azam Khalid, who has

provided many needed laughs, warm encouragements, and valuable advice.
















TABLE OF CONTENTS

A C K N O W L E D G M E N T S ......... ................................................................................... iii

L IST O F T A B L E S ................................................................... ..................... vii

FIGURES ............... ....... ...............................viii

ABSTRACT .............. ..................... ........... .............. ix

CHAPTER

1 ENVIRONMENTAL CONTEXT CHANGE EFFECT.................. ............... 1

H abitu action ........................................... 10
S tra te g y ............................................................................. 1 3

2 EXPERIMENT ........................................... .............. .. 19

Tim e and Strategy Predictions....................................................... ..... ....... .. 19
W working M em ory Predictions............................................................. .............. 2 1
Methods .................................... ...... .............. 22
Participants...................... ........ 23
M materials ......................................... 23
P procedure ........................................ 24
E ncoding ..................................... ............................ 24
R e call ........................................ 2 5
Post-recall measures............................... ................... 25
Results.................. .............................. .............. 26
Time, Strategy, and Environment Variables.............................. ................... 26
W working M em ory............................. .............. 27
R retrospective R reports ................................................................ ............. 30
Post Hoc Analyses ............. ........................... 31

3 GENERAL DISCUSSION ........................................ 34

Environm ent........................................ ........ 34
Encoding Strategy ..................................... ....................... 35
T im e ......................................................................................... 3 6
Working Memory.................................. .............. 36
Population Sample Problems .................................. .................... .... .... 38



v











G general C onclu sion ............... ...... .................... .. ...... ............. ................ 39

L IST O F R E FE R E N C E S ....................................................... ...................................... 4 1

BIOGRAPHICAL SKETCH .............................................................. ................. 47















LIST OF TABLES


Table Page

1-1. Recall Mean Proportions and Standard Deviations for Early, Late, Match, Mismatch,
Shallow Encoding, and Deep Encoding Conditions ............................................. 27

1-2. Recall Mean Proportions and Standard Deviations for Introductory Psychology
C course Students ................. ....... ..................................... ................. 32

1-3. Recall Mean Proportions and Standard Deviations for Upper-Division Psychology
Course Students ...................................... ............................... ........ 33



















FIGURES
Figure Page

1-1. Triangle C Span Task to Assess Working Memory Span Example Slides .............. 26

1-2. Best-fitting Regression Lines for Working Memory and Proportion of Words
Recalled in Each Environment for Shallow Encoding Condition............................. 29

1-3. Best-fitting Regression Lines for Working Memory and Proportion of Words
Recalled in Each Environment for Deep Encoding Condition............................... 30















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

INFLUENCE OF TIME SPENT
IN AN ENVIRONMENT AND ENCODING STRATEGIES
ON THE ENVIRONMENTAL CONTEXT CHANGE EFFECT

By

Khanh N. Nghiem

August 2006

Chair: Ira Fischler
Cochair: Peter F. Delaney
Major Department: Psychology

The environmental context change effect is the finding that forgetting occurs

when participants are tested in environments that do not match the encoding

environment. Regardless of the high frequency of anecdotes that are reported,

environmental context effects have been difficult to replicate or are modest. The

following study includes one experiment that examined two variables that have not been

extensively explored: time spent in the initial environment prior to studying and initial

encoding strategies. Participants' initial encoding strategy was fixed and some

participants studied the words immediately or after a delay. They were then asked to

recall the words in an environment that either matched the encoding environment or

mismatched. Participants who were required to do a shallow encoding strategy were

expected to have better recall when the environment matches, replicating the















environmental context change effect from past research, and to do better when studying

the words after a delay than when words were studied immediately. After a delay,

participants may be habituating to the environment and environmental cues will not be

encoded with the target items. Participants who were required to engage in a deep

encoding strategy were expected to not display any forgetting due to mismatch of

environments regardless of duration of time spent in an environment. Deep encoding

strategies allow for better retrieval cues in which there are inter-item associative

processing where one item primes the retrieval of another item. Working memory was

also measured as past studies have found that working memory is related to attentional

control. A 2 Time of Encoding x 2 Encoding Strategy x 2 Environment between-subjects

ANOVA with match and mismatch environments for the first factor, early and later

encoding times as the second factor, and encoding strategy as the third factor revealed no

significant main effects for time or significant interactions. The main effect of strategy

was significant; participants in the deep encoding strategy have higher recall than those in

the shallow encoding strategy regardless of time or if the environment matches or

mismatches. There were no significant correlations in working memory with any of the

variables of interest (time, strategy, environments). The results suggest that perhaps

encoding strategy plays a much stronger role in environmental context change effect than

was previously expected. However, problems with the population sample may have

resulted in undetectable effects.














CHAPTER 1
ENVIRONMENTAL CONTEXT CHANGE EFFECT

In the coming-of-age movie Cinema Paradiso, a famous filmmaker returns to his

small hometown after an absence of 30 years for his mentor's funeral. After

encountering his house and the dilapidated movie theater that was once the town's soul,

memories of his childhood with the man who inspired his dream to become a filmmaker

and memories of his first love come drifting up. The film is dramatically poignant in that

these memories are recalled with a haunting tinge of regret, nostalgia, and realization of

the joy and complexity of life.

Our everyday life plays against a backdrop of environmental and sensory

information. It is not surprising that our memories are intrinsically tied to contextually

rich information. These contextual cues have the power to evoke vivid and intense

memories. All of us have at one point or other experienced this-a song playing that

triggers a memory of a time we heard the tune, catching a whiff of a scent that reminds us

of a particular person, or returning to a previous place in the past and finding formerly

unavailable memories are dredged up. Our memories are rich with these sensory,

contextual cues.

The body of research on context is vast, examining such diverse contextual factors

as mood (Eich, 1980; 1985), internal context (Eich, 1975; 1980; Weingartner, Adefis, &

Eich, 1976), and environmental context (Smith, 1979; 1984; 1986; Smith, Glenberg, &

Bjork, 1978). Experimental context change paradigms have been used extensively to

study context reinstatement effects (Krafka & Penrod, 1985; Smith, 1979; 1984; 1985a),









type of study material (Vela's study as cited in Smith, 1988; Eich, 1985; Steuck & Levy's

study as cited in Smith, 1988), and type of environment (Godden & Baddeley, 1975;

1980; Saufley, Otaka, & Bavaresco, 1986; Smith, 1985a; 1985b). The environmental

context change study paradigm can be divided into two categories: first-order and

second-order experimental paradigms (Bjork & Richardson-Klavehn, 1988; Smith &

Vela, 2001). First order experimental paradigms involve participants studying material in

one environment and then being tested in the same environment (control condition) or a

different environment (context change condition). Godden and Baddeley's (1975) classic

study is an example of a first order paradigm that examined context change where

participants studied on land or under water and testing was either in the same

environment or the other environment (e.g. study on land, recall under water). First order

experimental paradigms provide a simple manipulation of context change and thus

criticism has arisen that participants may be able to mentally reinstate their environment.

Second order experimental paradigms involve participants studying at more than one

environment and with more than one list; hence it is more difficult for participants to

mentally reinstate their environment (Smith, 1984; Smith & Vela, 2001).

Environmental context examined in this experimental proposal can be specified in

the following dimensions: external (vs. internal), incidental (vs. deliberate), not focal (vs.

focal), and general (vs. specific) (Smith, 1988). Ostensibly, the distinction between

external and internal context may seem straightforward. However upon further

examination, external and internal context is more entwined than may be originally

assumed. External context is represented internally while internal context may alter the

perception of external context. However, for the sake of simplicity, researchers have









assumed that external context includes the physical aspect (i.e. environment) while

internal context includes any factors that can be represented within a person (i.e. mood,

thoughts, emotions). Environmental context is also incidental and not focal in that cues

that are not meaningfully related to the study itself-that is, it is outside the attentional

focus of the participant-are not actively utilized by the participant. Also, environmental

context is general; not any one specific aspect of the environment is manipulated but all

parts of the physical environment.

Of particular interest to this experimental paper is environmental context change.

We can all relate particular episodes in which forgetting occurred due to a change in

location: leaving a room to retrieve something only to forget it and suddenly

remembering again when returning to the original room or returning to a childhood house

and remembering certain incidents that were unlikely to be retrieved before. Regardless

of the high frequency of anecdotes that are reported, environmental context effects have

been difficult to replicate (Eich, 1985; Fernandez & Glenberg, 1985; Saufley, Otaka, &

Bavaresco, 1985) or produce small effects. Smith and Vela (2001) performed a meta-

analysis examining the reliability of context change and concluded that environmental

context effects were modest but reliable. The effect seems to be influenced by several

variables; of those identified by Smith and Vela were: degree of differences between

environments examined, time interval between learning and recalling, presence or

absence of the same experimenter, and type of material learned (i.e. associative or non-

associative).

Not surprisingly, the more the encoding environment matches the retrieval

environment, the smaller the magnitude of forgetting due to changing environments.









Smith and colleagues (1984, 1988, 2001) have suggested that the greater the resemblance

between the encoding and testing environments, the more likely participants will

mentally reinstate the original environment. Hence, the context change effect is reduced.

In additional, increasing the time between learning and recall will bolster the context

change effect. Accessibility of the memory traces may be weakened and hence more

dependency on cue reinstatement, as in matching environments. This results in greater

reliance on contextual cues and a larger context change effect is observed. The presence

of the same experimenter during both encoding and testing phases of the experiment for

the mismatch environment conditions would contaminate context cues for the new

environment. The experimenter is viewed as part of the environment and in a sense, can

serve as a retrieval cue. Therefore, when the same experimenter is present in the new

environment, this environment has both new and old aspects (never exposed and

previously exposed environments). Even more detrimental to experimental results would

be if the same experimenter is also present during the testing phase. Participants spend a

large amount of time during experiments interacting with the experimenter. Therefore,

the presence of the same experimenter after an environment change is a poor

experimental control. Consistent with this idea, Smith and Vela found that switching

experimenters for mismatch environment conditions resulted in increased context change

effects. Interestingly, the type of material studied can influence the magnitude of context

change effects. Materials that elicit inter-item associative processing in which one item

may potentially prime the other during retrieval were found to reduce the context change

effect. Presumably, this was due to less reliance on environmental cues due to better

retrieval cues.









Several hypotheses have been proposed to explain environmental context change

effects. One of the first explanations posited that the context change effect is dependent

on proactive interference. Context change that reduces proactive interference to benefit

memory can be illustrated by naturalistic evidence. Smith (1988) points out vacations

often facilitate creativity and productivity by aiding the vacationer in forgetting worries

and trite ideas (i.e. "writer's block," "stuck-in-a-rut" conditions). We often refer to

vacations as "escapes" and by going to an unfamiliar, novel environment, interest is

renewed. In problem solving studies, it has been suggested that incubation or setting

aside the problem when stuck and then later attempting to solve the problem may be

dependent on context change (Smith, 1995). Smith suggests that a context change

reduces fixation on and accessibility to wrong solutions. This idea could be observed by

our tendency to "take a break" when a task becomes too frustrating or tedious. Although

reduction of proactive interference may be a likely explanation for overall better

performance due to a change in context, it does not adequately explain the forgetting that

does occur when changing context. Early studies of environmental context suggested

that a context change lead to a reduction in proactive interference, which resulted in

better recall (Dallett & Wilcox, 1968). However, these studies included learning in

multiple contexts, which is a different paradigm than the one used in this study. Studying

lists of words in multiple contexts leads to better recall than studying in the same context

because the environment acts as an organizational cue (Smith, 1982; 1984). This finding

is supported by classroom studies in which no environmental context change effect was

observed because students usually studied in many locations outside the classroom

(Abernethy, 1940; Chen's study as cited in Smith, 1988). Reduction of proactive









interference as a plausible explanation of context change effect has not been widely

accepted. The context change effect refers to forgetting information due to context shifts

whereas release from proactive interference generally leads to better recall or less

forgetting (Wickens, 1970).

Physical disruption during the changing of contexts was also suggested as a

possibility for the context change effect. Strand (1970) found that a disruption during

encoding of words also display forgetting and posited physical disruption during

changing environments as driving the context change effect. After the word presentation,

participants were interrupted and asked to wait in the hallway. The time interval during

the wait was equal to the amount of time walking participants to a new environment for

the mismatch conditions. Participants waiting in the hallways (match conditions) were

returned to their original environments for recall. This explanation has been discredited

by studies that found reliable context change effects even after accounting for the

physical disruption (Smith, 1979; Smith, Glenberg, & Bjork, 1978). Subsequent studies

of context change effects included a physical disruption for all conditions to control for

this potential confounding.

Tulving's (1973) encoding specificity principle states that the match between

encoding and retrieval episodes is most important for facilitating successful recall. The

more similar the two episodes are in terms of intrinsic and extrinsic components, the

more likely it is that the memory trace is recalled. This encoding specificity principle

forms the underlying basis for the similarity principle, which provides a context change

effect explanation. According to the similarity principle, the context change effect is due

to the higher degree of difference between mismatched environments than when the









environments are the same. Ideally, environments with the most differences on the most

dimensions will result in a larger context change effect. However, this hypothesis is too

general: it does not specify what these intrinsic and extrinsic cues are. It is difficult to

make predictions, experimentally manipulate variables that will produce the most context

change, or control for variability because these cues are not defined.

The integration hypothesis (Baddeley, 1982) states that in order for the context to

influence memory the study material must be integrated with contextual cues during

encoding. Therefore, recall of the study material is highly dependent on these contextual

cues. This hypothesis has recently found support by studies examining context

integration (Earles, Smith, & Park, 1994; Eich, 1985; Park, Smith, Morrell, Puglisi, &

Dudley, 1990; Smith, Park, Earles, Shaw, & Whiting, 1998). These studies concluded

that environmental context cues were utilized only when participants were instructed to.

Not surprisingly, a context change effect was observed only when participants were

instructed to integrate the study materials with elements of the environment. However,

the integration process does not predict any context change effect for incidental

environmental context effects. Past studies have found reliable effects (Smith, 1979,

1985; 1986) and therefore this hypothesis has been viewed as not a strong explanation of

environmental context change effect.

Eich's (1995) mood mediation hypothesis states that environmental context change

can be explained by mood-dependent memory. Participants' moods vary across

environmental episodes and these moods become intrinsically tied to environmental cues.

Therefore, during retrieval, participants can access mood dependent cues as retrieval cues

to recall. Hence, when the learning and testing environments match, participants are able









to access their mood-dependent memory. A context change effect is observed when the

environments do not match and participants cannot access their mood-dependent

memory. However, mood mediation hypothesis has been unable to explain the effects of

material type eliciting associative or non-associative processing, test type, etc. (Smith,

1995; Smith & Vela, 2001). Researchers have pointed out that mood-dependent memory

may be a reflection of mental context (Smith, 1995). It is unlikely that changing

locations could produce a large enough change in mood that would be robust; even strong

effort to manipulate mood directly rarely result in effect sizes that reach or exceed the

environmental context change effects.

As an alternative, Glenberg (1997) suggested that information from the world is

continuously represented unless suppressed, which occurs in order to free up cognitive

resources for more demanding tasks. In other words, environmental context cues are

suppressed due to cognitive load. This idea informed two explanations of environmental

context-dependent memory: the overshadowing hypothesis and the outshining

hypothesis.

The overshadowing hypothesis (as described in Smith, 1988; Smith & Vela. 2001)

states that context changes are due to a failure to encode environment contextual cues for

a variety of reasons, such as control of attention or conceptual processing of study

materials. Like the term implies, if environmental cues are overshadowed by other

factors during encoding and are not stored, contextual change will not affect recall.

Therefore, in the absence of factors that may overshadow environmental cues (e.g.

material that allows for associative processing or that draws attention away from the

environment), a context change should be observed.









Smith's (1978; 1979; 1985; 1988) outshining hypothesis proposes that

environmental cues are not used if other better cues are available. In other words, better

non-contextual cues will outshine environmental cues much like the sun outshining all

other celestial bodies in our sky. Unlike the overshadowing hypothesis, the outshining

hypothesis states that environmental context cues are encoded, but are not utilized during

retrieval because better cues are available. Better cues can be defined as deeper

processing of the material, generation of better retrieval cues due to deeper encoding

strategies, or those that encourage inter-item associations.

The overshadowing and the outshining hypotheses may seem similar in that they

both specifies failures of the environmental context change effect, but they are actually

different in their predictions of context change. One distinction is that the overshadowing

hypothesis is concerned with failure to store contextual cues at encoding and therefore

there are none to use at retrieval. The outshining hypothesis suggests that contextual cues

may be encoded successfully but the cues are not used at retrieval because better retrieval

cues are available. Both are dependent on associative processing and inter-item

associations; however, overshadowing specifically explains context change effects as

encoding effects while outshining specifically explains them as retrieval effects.

These theories have been largely unsuccessful at adequately predicting why and

how context change effect occurs. Smith and Vela's meta-analysis resulted in mixed

findings for the overshadowing hypothesis and outshining hypothesis. It could be that

these studies are not exploring variables that may influence encoding and retrieval-

hence, affecting the magnitude of the context change effect. This may explain the mixed

findings in the environmental context change literature. Therefore, I am proposing two









variables that may influence the context change effect and should be empirically

investigated: habituation and strategy.

Habituation

How attention fluctuates over time is important in examining environmental

context change. It has been shown that participants tend to mentally reinstate their

environments by imagining the initial learning environment (Smith, 1979; 1984),

resulting in a decrease in context change effect. In the absence of such a strategy, context

change effects are observed. Glenberg, Schroeder, and Robertson (1998) found that

when participants' gazes were diverted from the environment during retrieval, recall was

better due to less processing devoted to the present environment and more so to

conceptual processing. Perhaps attentional control is related to effectively processing and

retrieving environmental cues.

In Glenberg's (1997) theory of environmental suppression, a shifting of cognitive

resources must occur for conceptual processing. These cognitive resources seem to be

attentional resources. Indeed, the overshadowing hypothesis and outshining hypothesis

both imply that if attention is drawn away from the learning or testing environment, no

environment contextual manipulations should be observed. It would seem intuitive to

state that most people are aware of their environment, but that the attention fluctuates

from the environment to the task, or focal information. The role of attention in context

change studies seems to be under-explored and not examined thoroughly although most

researchers agree that attention is important. Experimental manipulations to make the

environmental cues more prevalent (Godden & Baddeley, 1975), novel or flashy (Dallet

& Wilcox. 1968; Dulsky, 1935; Weiss & Margolius, 1954), or included explicit

instructions to note the environment (Eich, 1985; Nixon & Kanak, 1981) have all been









aimed to capture or direct attention. The environmental contexts for Godden and

Baddeley's study were under water and on land. Participants in Dallet and Wilcox's

study were asked to wear a box of changing lights over their heads rather than change

rooms. The box not only succeeded in producing a context change effect, but also caused

nausea and disorientation for the participants. Most of us do not spend a large amount of

our time in water, wearing scuba gear, or adorning our heads with a box full of lights.

The box of changing lights and the equipment for going under water (scuba diving gear)

as well as aspects of the environment (physical suspension in water, physiological

changes) were distinct and novel enough that attention was more engaged to the

environment than usual. These experimental manipulations contained unusual elements

that drew attention.

Environments oftentimes are familiar and frequented: we work at a certain place,

have a home at a certain location, or visit preferred stores. On the other hand when an

individual first encounters or enters a new environment, the environment is unfamiliar

and novel; thus the individual attends or orients to it. After a length of exposure to this

environment, the orienting response is reduced and habituation occurs (Cowan, 1988;

Groves & Thompson, 1970; Sokolov, 1975). Dehabituation or revival of the orienting

response may appear if aspects of the environment capture attention or attention are

directed to the environment. The longer the participant spends in an environment, the

less novel and interesting the environment is. For example, waiting in the lobby at a new

doctor's office, I swiftly surveyed the furniture arrangement and even find myself

studying the prints hung against cheery wallpaper. As time went by (and it was obvious









my doctor was running late), I was no longer interested in the features of the room but

instead perused the magazines available.

Habituation occurs over time and predicts that the longer the time spent in an

environment, the less likely environmental cues will be utilized. Therefore, when

participants are asked to study information, the environmental cues are less likely to be

incidentally encoded with the study information. If the participant immediately studies

information upon entering the environment, the participant's attentional resources may

still be allocated to perceiving the environment. Therefore it would seem that

participants who study information immediately would also incidentally encode

environmental cues. When the environment at recall does not match the one at studying,

these participants are more likely to show the phenomenon of forgetting due to mismatch

of environments. Using the example above, if I was asked to remember a list of

questions I wanted to ask my doctor during the time that I was still actively interested in

the room, I would be more apt to forget the questions once I entered the doctor's

examination room. Upon entering the lobby again, I belatedly remembered the questions.

If, however, I was asked to remember the questions after I was no longer actively

interested in my physical environment, I would be more likely to remember them upon

entering the doctor's examination room. Less forgetting due to changing environments

occurs because attention is no longer focused on the environment.

Unfortunately, the literature on environmental context change has not extensively

examined amount of time spent in an environment. Most real world events occur in a

particular environment for more than the mere couple of minutes that are examined in

laboratory studies. It would be more ecologically valid to examine the influence of time









in an environment. A recent study (Isarida, 2005) found that the longer time spent

studying the words, the greater the magnitude of contextual information integrated with

the study materials. Isarida proposed that environmental context effect is related to

amount of time spent studying. This account is different from the habituation view in

that Isarida predicts that time during encoding will increase the associative strength

between the context and the item. The habituation view proposes that attention drawn to

the environment before and during encoding results in sampling of the environment

which will be incidentally encoded with the study materials.

According to the overshadowing hypothesis, the longer the time spent in the

learning environment, the higher the chances of the contextual cues incidentally encoded

with the studied materials. Smith and Vela's meta-analysis did not show amount of time

spent in an encoding environment influencing recall performance. Effect sizes were

calculated and compared for studies that had examined environmental dependent

memory.

But these studies did not include amount of time spent in the environment prior to

learning. Examined was time spent in the learning environment either during learning,

such as completing a distractor task in between presentation of words or after learning,

such as completing a distractor task while still in the learning environment. No studies

have yet looked at time spent in an environment prior to studying.

Strategy

Another variable that has not been extensive examined and may affect the

environmental context change effect is participant's initial encoding strategy. In their

meta-analysis, Smith and Vela found that study materials that do not induce associative

processing showed a greater context change effect whereas associative processing









resulted in much smaller context dependency effects. These studies range from

examining depth of processing (Murnane & Phelps, 1995; Smith, 1986; Smith, Vela, &

Williamson, 1988) to varying the strength of retrieval cues of the study materials by

associative processing. For example, in Smith's (1986) study, the shallow processing

task was actually an incidental short term memory task in which participants heard a

short list of words and after several seconds, immediately recalled them or a "deeper"

processing task in which participants were told to attempt to memorize the words.

Afterwards, participants completed a variety of distractor tasks and then either recalled in

the original or new environment. A context change effect was observed for the shallow

processing but not for the deeper processing.

Studies that have looked at inter-item associations and associative processing of

study materials found that these reduced or eliminated the context change effect. Smith

and Vela (2001) suggested that participants do not utilize contextual cues even if they are

available because associative processing provides better retrieval cues: Recalling one

item would guide the retrieval of another. This outcome can be explained by both the

overshadowing and outshining hypothesis, but both attribute it to different reasons. The

overshadowing hypothesis predicts that associative processing draws attention away from

the environment and therefore environmental cues are not encoded and therefore are not

available for use. The outshining hypothesis predicts that environmental cues do get

encoded but associative processing outshines the environmental cues as retrieval cues and

are, thus, utilized.

Einstein and Hunt (1980) examined the levels of processing and organizational

approach to encoding. They point out that levels of processing tasks often involve









processing individual items whereas organizational strategies involve processing between

items. They conclude that there is a distinction between individual item based processing

and relational processing and that this distinction is important in regards to defining

elaborative processing. Craik and Lockhart's (1972) concept of elaborative processing

involves individual item processing where items are made more meaningful and hence

more accessible in memory. According to the levels of processing perspective, rote

rehearsal is viewed as shallow processing because items are repeated in short term

memory. Elaborate processing is rehearsing the items to store them into long term

memory. The levels of processing view would argue that elaborative processing would

result in better recall. However, Benjamin and Bjork (2000) found that in the presence of

a time pressure, accessibility to items was more disrupted for elaborative rehearsal than

rote rehearsal. They concluded that elaborative rehearsal may facilitate stronger retrieval

cues but there is a time trade-off. Elaborate rehearsal is effortful and resource

consuming; engaging in elaborate rehearsal during encoding and resurrecting the mental

framework during retrieval is time consuming (Masson & McDaniel, 1981). However,

there is evidence suggesting that the mere time and effort of elaborative processing of

individual items is not enough to strengthen accessibility cues. Bradley and Glenberg

(1983) found that time and attention spent during rehearsing items individually did not

enhance recall. There was enhanced recall only when more than one item are rehearsed

together, forming and strengthening inter-item associations.

Environmental context studies have only manipulated processing in the Lockhart

and Craik sense. These studies utilizing associative processing have found that this

greatly reduces or eliminates context dependency. However, no studies have looked at









encoding strategies that require relational item processing. In contrast to the Lockhart

and Craik levels of processing view, depth of encoding strategies involve both forming

inter-item associations as well as generating individual item distinctiveness. Therefore,

shallow encoding would be viewed as maintenance rehearsal of individual items or

rhyming the words. One form of deep encoding involves generating an interactive image

or a story for all the items (Bower, Clark, Lesgold, & Winzenz, 1969). This

categorization of encoding strategies has been utilized by researchers examining

recollection (Perfect & Dasgupta, 1997), spacing effects (Delaney & Knowles, 2005),

and directed forgetting (Sahakyan & Delaney, 2003; Sahakyan, Delaney, & Kelley,

2004).

Sahakyan and Delaney (2003) have found that changing internal contexts induces

strategy changes in directed forgetting studies. In directed forgetting studies, participants

are presented with two word lists and half are told to intentionally forget the first list. A

typical outcome is the presence of both costs and benefits of directed forgetting, in which

costs refer to the poorer recall that forget participants display for the first list relative to

remember conditions and benefits refer to the increased recall that forget participants

display for the second list relative to the remember conditions. In a series of studies,

Sahakyan and Kelley (2002) and Sahakyan (2004) have argued that the costs in directed

forgetting can be attributed to a mental context change. When instructed to forget the

list, participants attempt to "think of something else", which then changes their mental

context. Therefore the context during encoding of the first list and the context during

recall are mismatched, leading to poorer recall for the first list or costs. The benefits,

however, are due to strategy changes that are induced by the mental context change.









Sahakyan and colleagues (Sahakyan & Delaney, 2003; Sahakyan, Delaney, & Kelley,

2004) have suggested that the mental context change that forget participants engage in

are also evaluative in nature. Participants evaluate their encoding strategy for list one

after being told to forget it and perhaps realize that their current method is ineffective,

hence a strategy change occurs for the second list. Without prior knowledge or training,

participants tend to engage in shallow encoding, such as rote rehearsal (Delaney &

Knowles, 2005; Sahakyan & Delaney, 2003; Sahakyan, Delaney, & Kelley, 2004).

Therefore switching to a deeper encoding strategy would explain the increase in recall for

the second list for forget participants or the observed benefits. They concluded that

utilization of a deep encoding strategy, especially those that initially engaged in shallow

encoding, can improve recall due to better encoding of the information and better

retrieval cues.

Deep encoding contains both relational and item-specific processing; generating

an interactive image or story allows for each item to become linked or associated to

another item. Retrieval of an item would therefore guide recall of another item. Based

on the above research findings, the type of strategy during encoding may account for

context dependency. Deep encoding may provide resistance to forgetting due to

changing contexts. Because the majority of individuals without prior training engage in

shallow encoding, the context change effect is observed. If instead deep encoding was

used, the context change effect should be reduced or eliminated since associations

between items serve as better retrieval cues than environmental cues. This predication

would seem similar to the outshining hypothesis in that there are better retrieval cues

available and hence environmental cues are outshone, however the outshining hypothesis









predicts that environmental cues are successfully encoded. Engaging in an elaborative

processing task may potentially interfere with encoding of environmental cues as

cognitive resources are allocated to the task, a suggestion of the overshadowing

hypothesis. However, the overshadowing hypothesis attributed the context change effect

to this. Use of a shallow strategy is not as effortful and will not detract from encoding of

environmental cues and use of a deep strategy is more effortful and may detract from

encoding of the cues. However, the lack of environmental cues is not necessary since the

deep strategy provides better retrieval cues than environmental context cues. It would be

interesting to examine further these ideas as no studies in the environmental context

literature, thus far, have extensively examined relational encoding strategies.














CHAPTER 2
EXPERIMENT

The purpose of the first experiment was to examine the effect of time spent in the

encoding environment and recall of words in matched or mismatched environments while

controlling for participants' encoding strategies. Participants that utilize deeper

processing such as inter-item associations usually display resistance to contextual

manipulations because recall of one item primes the recall of another item (Smith, 1988;

Smith & Vela, 2001; Steuck & Levy's study as cited in Smith, 1988). In this sense, deep

encoding of the material may result in items becoming more associated with one another

and more meaningful in memory. When information to be studied is more meaningful,

environmental context effects have been observed to be lower than when the study

materials are meaningful (Steuck & Levy's study as cited in Smith, 1988; Vela's study as

cited in Smith, 1988). In light of this, experimentally controlling participants' encoding

strategies would allow closer examination of the separate influences of time and

participants' encoding strategies. I expect to find evidence supporting the following

predictions of the influence of habituation and deep encoding strategies on the magnitude

of the environmental context change effect.

Time and Strategy Predictions

Overall, participants in the match conditions are expected to do better than those in

the mismatch conditions, replicating the environmental context change effect. This

finding will replicate studies that found reliable context-change effect (see Smith & Vela,

2001) in that there is forgetting that occurs when changing to a physical environment









different from that for encoding. Participants who study the words immediately in the

experiment should have poorer recall than participants who waited seven minutes and

then studied the words. This may be because participants who immediately studied the

words may attenuate to the environment. In other words, those that study the words after

a delay have habituated to the environment. Hence, they have more attentional resources

to allocate to the task than those that have not habituated to the environment and are still

attenuating.

Participants trained through a deep encoding strategy are expected to have higher

recall than those trained through a shallow encoding strategy regardless of influence of

time or if the environment matches. Participants who were instructed through a shallow

encoding strategy should be affected by time spent in an environment prior to studying

the words. In particular, recall should be worst for those who studied the words

immediately than those who studied the words after a delay when the environments

mismatches, indicating a context change effect. For match conditions, there should be no

influence of time since environmental cues are available for retrieval.

There should also be a significant recall environment by times of encoding

interaction. Participants that study words immediately (early condition) in environments

that match for encoding and recall should have the highest recall. The conditions that are

expected to display the worst recall are the ones in which participants study the words

immediately and then change environments (late and mismatch environment condition).

For participants who study the words after waiting (the late condition), those that are in

environments that match for encoding and recall should do better than those that are in

environments that mismatch. This finding will support what was earlier predicted:









Habituation to the environment reduces the forgetting that occurs when environments

mismatch. Participants who study words immediately incidentally encode environmental

cues with the target words, resulting in a dependence on environmental context cues.

Therefore when the recall environment is different from the initial environment in which

the words were studied, recall should be reduced.

Working Memory Predictions

Working memory was also included as a measure due to its role in attentional

control (Kane & Engle, 2000). Delaney and Sahakyan (2004) demonstrated that mental

context change is related to working memory capacity. They instructed participants to

continue to remember a list of words and then a mental context change task was

administered. They found that high working memory participants were more affected by

a mental context change, forgetting more information than low span participants. Based

on these findings, I expect individual differences in working memory span to be related

to recall in conditions where there is a change of environment but not in conditions where

there is no environment change. In particular, high span working memory individuals'

recall may be reduced when there is a physical environmental change than low span

individuals. If this finding was supported, consistent with Delaney and Sahakyan's

(2004) results, high span working memory individuals would, thus, be more context-

dependent than low span individuals.

Working memory is expected to significantly correlate with recall from

experimental conditions (i.e., strategy, time and environment manipulations). Those with

higher working memory spans will be more likely to spontaneously use deep encoding

strategies. Encoding strategies are effortful and maintaining these processes after an

environment has changed may require the participant to inhibit attenuating to the new









environment and resurrect the mental scaffold of the effortful encoding strategy during

recall. Individuals with high working memory span should have decreased performance

when environments are changed. However individuals with relatively low working

memory span should remain unaffected by the environment changes. This may be

because those with low working memory span may not need to inhibit the environment

while engaged in effortful encoding strategies. Time is not expected to have a significant

correlation with working memory. There is no indication from the research literature that

suggests that habituation is influenced by working memory span. Therefore, the effects

of the length of time spent in the encoding environment should probably be unaffected by

working memory span.

Methods

In this experiment, the following variables were manipulated: initial encoding

strategies, duration of time spent in an environment prior to learning, and whether the

environment for learning and recall matched or mismatched. It has been argued that

salient elements of the initial environment may still be present in the second environment,

drastically reducing the environmental context change effect (Fernandez & Glenberg,

1985; Smith & Vela, 2001). For instance, the laboratory room for this environment is

very similar to a lobby or office; both include similar office furniture and arrangements or

similar presentation instruments. Therefore, to maximize contextual change, care is taken

to ensure no part of the initial environment is salient when changing environments. In

this experimental proposal, environmental context change is manipulated when the

environment is changed from an outside to inside location and vice versa, different









experimenters for each phase of the change, and different methods of studying the words

(on a multimedia player on the computer versus on a portable stereo).

Delaney and Knowles (2005) found that for participants engaged in shallow

encoding, the most frequent strategy reported was verbal rehearsal and for deep encoding

was a story generation mnemonic. In Delaney and Knowles's Experiment 2, they

experimentally controlled participants' encoding strategies by training participants

through a shallow or deep encoding strategy and requiring the usage of this strategy

throughout the experiment. A similar method was used to fix participants' initial

encoding strategy in this experiment.

Participants

Participants were University of Florida undergraduates who received course credit

or extra credit for completing the experiment. Participants were tested individually with

8 participants in each of the eight experimental conditions for a total of 64 participants.

Materials

Twenty unrelated, medium-frequency English nouns were selected. The words

were recorded by a male voice at a rate of 1 word per 4 s and edited on Adobe Audition

1.5. The inside environment is a computer lab room inside the psychology building with

a computer, two chairs, file cabinets and shelves, various computer equipment on the

ground and shelves, books and stacks of paper on the shelves, and white walls. The

outside environment is a relatively secluded and quiet picnic place outside the

psychology building. This picnic place is surrounded by trees, has a roof over it and

includes a stone table and stone benches on two sides of the table. The words were

presented on a multimedia player on the computer for the inside condition. For the

outside condition, the words were presented on a portable stereo with a compact disc









player. For the purposes of this experiment, different experimenters were considered

different for only one variable: gender. Gender was considered a substantial enough

difference for conditions where the environment was changed. Therefore, the gender of

experimenters was counterbalanced to control for possible biases. Also, care was taken

to ensure that the experimenters' appearance were not similar (e.g. clothes, hairstyle).

Procedure

Encoding

Participants were tested individually. Before they began the experiment, informed

consent was obtained from the participant in the hallway or neutral zone. Then they were

led to the encoding environment and asked to sit facing away from the experimenter to

discourage focusing solely on the experimenter. Half of the participants were told the

experimenter was still setting up the experiment (the late condition). They were told to

just sit quietly and wait until the experimenter finished preparing the experiment. The

experimenter was actually surfing the internet on the computer for the inside condition or

engaged in scoring data or a search-a-word game for the outside condition to appear

occupied. The participant was discouraged from studying or from engaging in

conversation with the experimenter to give the participant an opportunity to observe the

environment. After seven minutes, the participant was told that the experiment was ready

to begin and was then trained through a shallow or deep encoding strategy. The other

half of the participants, upon entering the encoding environment, was immediately asked

if they were ready to begin the experiment and trained through either a shallow or deep

encoding strategy. Participants in the shallow condition were instructed in the verbal

rehearsal encoding strategy prior to studying the words. They were instructed to rehearse

the words out loud, adding each new word to the set already rehearsed. They were also









told not to be alarmed if they forgot any words and to continue rehearsing the words that

they do remember. Participants in the deep encoding condition were instructed in the

story generation encoding strategy where they were required to generate aloud a story

utilizing each word. They were also told that they are required to use all the words in the

story. After studying the words, participants engaged in a math distractor task for 90 s to

reduce recency effects.

Recall

Participants in the match environment condition were led to the neutral zone and

then were led back to the original environment by the same experimenter. For the

mismatch environment condition, participants were led to the outside or the inside

environment depending on their initial encoding environment and a different

experimenter conducted the next portion of the experiment. Time between changing

environments was controlled for both match and mismatch conditions so that the time for

walking the participant to the new environment and the time for walking the participant

to a neutral zone and back to the same environment is the same: 90 seconds. They were

then asked to free recall the words on a blank sheet of paper for 90 seconds.

Post-recall measures

Retrospective verbal reports were collected to determine participants' initial

affective states and whether it changed and what they were thinking for those in the late

condition. They were also asked if the strategy task was difficult (i.e. following

instructions or engaging in the task) and if the words were spoken clearly and audibly.

After this, all participants completed a working memory span task called the Triangle C

Span Task. The Triangle C Span Task involves viewing slides on Microsoft PowerPoint

that consist of light purple triangles, dark purple triangles, and light purple squares. Their










task is to count the light purple triangles out loud and remember this total. They will

view 2 6 slides before being asked to recall the sequence of total number of light purple

triangles on each slide. Triangle C Span scores range from 0 58. Figure 1-1 illustrates

example slides.




A A A

A A
*
0 A A
A A
A A


(a) (b) (c)


Figure 1-1. Triangle C Span Task to Assess Working Memory Span Example Slides.


Participants should count and remember 4 purple triangles in the (a) first slide and

8 purple triangles in the (b) second slide. When the prompt to recall the total number of

purple triangles from each slide appears (c), they should recall 4 and 8.

Results

Time, Strategy, and Environment Variables

In order to ensure proper counterbalancing, variables such as gender of

experimenters and environment order were included in all main analyses; but since there

are no significant interactions, were collapsed over. Mean proportions of words were

calculated. A 2 Time of Encoding x 2 Encoding Strategy x 2 Environment between-

subjects ANOVA with match and mismatch environments for the first factor, early and

later encoding times as the second factor, and encoding strategy as the third factor

revealed no significant main effects for time F(1, 56) =.008, MSE= .0002, p = .929,









environment F(1, 56) = 2.031, MSE =.040, p = .160, or significant interactions: for time

and strategy, F(1, 56) = .0006, MSE = .032, p = .859; for time and environment, F(1, 56)

= 2.570, MSE =.0510, p = .115; for strategy and environment, F(1, 56) = 1.785, MSE =

.0351, p = .187; for time and strategy and environment, F(1, 56) = .198, MSE= .0039, p

=.658. The main effect of strategy was significant, F(1, 56) = 28.555,MSE =.563,p <

.001. Participants in the deep encoding strategy have higher recall than those in the

shallow encoding strategy (see Table 1-1 for overall means and standard deviations)

regardless of time or if the environment matches or mismatches.

Table 1-1. Recall Mean Proportions and Standard Deviations for Early, Late, Match,
Mismatch, Shallow Encoding, and Deep Encoding Conditions



Strategy

Shallow Deep

Time Mean S.D. Mean S.D.

Early
Match .306 .073 .519 .189

Mismatch .343 .018 .494 .201

Late
Match .344 .132 .600 .136

Mismatch .300 .046 .431 .194



Working Memory

Nine participants did not have working memory span scores because their scores

were excluded due to computer failures or failure to follow instructions. These nine

participants were replaced. Participants' Triangle C Span scores ranged from 9 to 45,









with a mean of 25.42 (SD = 9.307). There were no significant correlations between

working memory and recall of the words, r = .002, p = .988, which was not what was

predicted.

To investigate the possibility that recall in the experimental conditions was

dependent on individual differences in working memory, linear regression analyses were

employed (Jaccard and Turrisi, 2003). Linear regression has been utilized by Delaney

and Sahakyan (2004) to examine individual differences in working memory for mental

context change effect. Linear regression analyses were conducted to examine proportion

of words recalled as a function of working memory, experimental conditions (time,

environments), and their interaction. Experimental condition was entered as orthogonal

contrast-coded variables and the analyses were conducted separately for each strategy

group (shallow and deep). The main effects were entered simultaneously and the

interaction terms were entered to determine if any of the interactions explain any

additional variance. The total model was significant for shallow encoding, F(6, 25) =

2.940,p < .05 but not for deep encoding, F(6, 25) = .794,p = .583. For shallow

encoding, there was a significant main effect of environment, F(1, 25) = 2.805, p < .01

and revealed a significant working memory and time interaction, F(1, 25) = 2.340, p <

.05. Thus, the effect of experimental conditions was modulated by working memory span

when participants were required to engage in a shallow encoding strategy. For shallow

encoders, the higher the working memory, the higher the proportion of words recalled for

matching; whereas for mismatch environments, differences in working memory span

does not influence recall environments (see Figure 1-2).












Strategy: shallow


0.60-






0.50-






O 0.40-






0.30-


Condition
O match
0 mismatch

















R Sq Linear = 0.316

R Sq Linear = 0.03


I I I I I
10 20 30 40 50
WM


Figure 1-2. Best-fitting Regression Lines for Working Memory and Proportion of Words
Recalled in Each Environment for Shallow Encoding Condition


For deep encoders, working memory is not influenced by physical environmental

manipulations (see Figure 1-3).


0~0 0 0


0 0 0 0


0.20-


0


0


00














Strategy: deep


Condition
O match
0 mismatch


















R Sq Linear = 0.022

R Sq Linear = 0.076


Figure 1-3. Best-fitting Regression Lines for Working Memory and Proportion of Words
Recalled in Each Environment for Deep Encoding Condition


Retrospective Reports

For the retrospective reports, participants were asked about their affective states

before the experiment, after the delay (for those in the late condition), and what they

were thinking about during the delay. These are experimental checks that ensure that

participants' recall is not better or poorer due to high arousal. Past research on emotion

and arousal has shown that higher arousal tends to result in better memory performance

(Anderson, 1988; Duffy, 1962; Mandler, 1975; Yerkes & Dodson, 1908). In this


0.90-



0.80-



0.70-



0.60-



0.50-



0.40-



0.30-



0.20-


0



0 0

0

0 0 00








O 00
0 0 0 0




00 0O

0

00 0

0


WM









experiment, participants' responses were categorized as either high arousal (e.g. excited,

impatient, tense, nervous, anxious), neutral (e.g. "feels fine", good, curious, relaxed,

calm, "normal"), or low arousal (e.g. sad, tired, sleepy, bored). When the experiment

started, 35.9% participants experienced high arousal, 48.4% experienced neutral arousal,

and 12.5% experienced low arousal. The affective states did not significantly affect

recall; in other words, despite the past findings on arousal and better memory

performance, participants' performed at a comparable rate whether they were highly

aroused, neutral, or had reduced arousal. Of those that were in the late condition, 53.1%

had different affective states before and after the delay (52.94% changed to a high arousal

state, 23.53% changed to a neutral arousal state, and 23.53% changed to a low arousal

state). However, this switching did not significantly affect recall. Although participants

found engaging in the strategy task was slightly difficult, this was more due to the length

of the word list when adding the words to the cycle of repetitions rather than any

difficulties with following the instructions. No participants reported having any

difficulties hearing the words clearly.

Post Hoc Analyses

Of the 64 participants that participated in the experiment, 48.4% (n = 31) were

from introductory psychology courses. The others, 51.6% (n = 33), were from upper

division psychology courses. This may have biased the results; to test this post hoc,

descriptive statistics for the data was analyzed separately and the mean proportions of

recall were compared for those in the introductory psychology courses and those in upper

division psychology courses. There were 2-6 of each type of student in each condition

cell; therefore, because the cells do not contain equal sample numbers, the data should be

interpreted with caution. However, the upper division psychology students had overall









slightly higher recall for both shallow and deep encoding than introductory psychology

students. This is especially true for deep encoding for some of the conditions, such as

early and match, M= .388 (SD = .085) for introductory students versus M= .650 (SD =

.173) for upper division students (see Table 1-2 and Table 1-3).

Tablel-2. Recall Mean Proportions and Standard Deviations for Introductory Psychology
Course Students



Strategy

Shallow Deep

Time Mean S.D. Mean S.D.

Early
Match .250 .050 .388 .085

Mismatch .342 .020 .450 .354

Late
Match .276 .035 .608 .139

Mismatch .300 .000 .438 .144











. Tablel-3. Recall Mean Proportions and Standard Deviations for Upper-Division
Psychology Course Students



Strategy

Shallow Deep

Time Mean S.D. Mean S.D.

Early
Match .340 .065 .650 .173

Mismatch .350 .000 .508 .174

Late
Match .367 .147 .625 .160

Mismatch .300 .071 .475 .203














CHAPTER 3
GENERAL DISCUSSION

Environment

Past research has shown that context change effect is a modest, although reliable

effect (Smith & Vela, 2001). However, no significant effect of environment was

observed; participants did equally well if the encoding and retrieval environments

matched or mismatched. Ferdnandez and Glenberg (1985) did not find any reliable

context change effects and actually observed a different context advantage as frequently

as not finding any advantage. They suggested that differences in environments and type

of disruption tasks in the literature may result in discrepancies in the magnitude of

context change effect. Perhaps the context change effect does occur but is undetectable

by current experimental manipulations.

This explanation, however, seems unlikely. Other researchers have been able to

find context change effects (Smith, 1979; 1984; 1986; Smith, Glenberg, & Bjork, 1978).

The environmental context change effect is also vulnerable to a number of different

manipulations, as was discussed in Smith and Vela's meta-analysis. In this study,

participants in the match condition were asked to walk to a neutral zone and then back to

the initial environment. In several of Smith's studies (1979, 1984, 1986), the participants

in the match condition were asked to wait in a neutral zone. This difference in disruption

task may account for no significant context change effects in the present study: Smith's

participants, while waiting in the neutral zone, may view this as an opportunity to

rehearse the studied words. Whereas participants in this study were continuously









walking for the duration of the disruption task and may view this task as beginning

another unrelated task. Therefore these participants are less likely to engage in rehearsal.

A criticism of past studies that did not find environmental context change effects

was that the environments were not different enough. Care was taken in this study to

maximize the differences between the two environments by choosing an indoor office

setting and an outdoor picnic table setting and the experimenters conducting the

environments. However, participants could be viewing the environments as part of the

same task (studying words) and therefore, their expectations did not alter. Manipulating

participants' expectations is a difficult experimental control. Isarida (2005) found that

type of task can contribute to context change effect, yet environmental context change

effect has been demonstrated for same-task studies (see introduction).

Encoding Strategy

There was a significant effect of strategy on recall. Participants that were

instructed to engage in a deep encoding strategy had higher recall than those that were

instructed through a shallow encoding strategy. Past environmental context research

found that deep processing resulted in better recall (Murnane & Phelps, 1995; Smith,

1986; Smith, Vela, & Williamson, 1988) possibly due to relational inter-item processing.

Therefore, an associative inter-item processing that converts the target items into more

meaningful information will result in better recall, as was observed in this study. This

replicates past research findings where deep encoding strategies result in better recall

than shallow encoding strategies (Delaney & Knowles, 2005; Sahakyan & Delaney,

2003; Sahakyan, Delaney, & Kelley, 2004). It may be that initial encoding strategy may

be powerful enough to eliminate any environmental or time influences. If this was the









case, then it would be interesting to examine shallow encoding strategy for future

research as the majority of participants engage in shallow encoding.

Time

It would seem from real-life observations that the more time spent in an

environment, the easier and better developed the representation of the environment will

be. Smith's overshadowing hypothesis (as described in Smith, 1994; Smith and Vela,

2001) actually predicts that the longer the time spent in an environment, the higher the

chance that environmental retrieval cues will be utilized. Therefore, changing

environments will result in poorer recall because of the utilization and greater

dependency of environmental cues. Regardless, time spent in an environment, whether

enhancing or reducing opportunities to utilize environmental context cues, seems

intuitively to significantly influence the magnitude of context change effect.

Surprisingly, there were no significant time manipulation effects; there were no support

for the habituation hypothesis as an adequate explanation of the context change effect.

One possible explanation may be that the time interval in this study (seven minutes)

was not enough for participants to habituate to the environment. Perhaps participants did

not attenuate to the environments as was hoped and had "zoned out" during the delay. It

does not seem to be that the environments must be radical to catch participants' attention;

however, it would seem that participants are more focused on completing the experiment

and therefore their attention becomes tunnel-vision.

Working Memory

There was mix support for the working memory predictions. There were no

significant correlations for working memory span and word recall, suggesting that

perhaps the environmental context change effect is not related to working memory. This









does not provide direct support for the prediction that individual differences in working

memory should also reflect and predict the magnitude of environmental context change

effect, specifically, that high span working memory individuals should be more affected

by changing environments. Delaney and Sahakyan (in press) found that when

participants were asked to forget information, high span individuals had poorer recall

compared to low span individuals. They attributed this to the mental context change that

occurs with a "forget" cue; high span individuals were more affected by the mental

context change than low span individuals. Surprisingly, this was not observed with

regular correlations as was predicted.

However, multiple linear regression analyses revealed interesting trends. Based on

the results of the linear regression analyses to assess individual differences in working

memory affecting recall of experimental conditions, it would seem that there was some

support for the hypotheses. For those engaged in a shallow encoding strategy, those with

high working memory span were more affected by the manipulations of environment and

time, resulting in lower recall when environments matched than mismatched, replicating

Delaney and Sahakyan's (2004) findings. For those engaged in a deep encoding strategy,

their individual differences in working memory did not influence experimental

manipulations, and hence did not affect recall. These findings provide some support for

the prediction that environmental and time manipulations should affect those in the

shallow encoding condition but not for those in the deep encoding condition.

Engle and Turner (1986, 1989) argued that working memory is independent of

task-related processing. However, it would seem more likely that deep encoding

provides such powerful retrieval cues that there is no influence of individual differences









in working memory span. Kane, et al. (2000, 2001) argued that working memory

capacity was attentional based. Due to the mixed working memory results, this also

suggests that perhaps environmental context change effect is less dependent on

attentional processing and is driven much more by strategy than was previously thought.

However, the trends are in the predicted direction.

Population Sample Problems

The retrospective reports showed that participants tend to become increasingly

more aroused after a delay (e.g. impatient, tense, nervous, excited). However, most

participants initially begin the experiment as either aroused or neutral (e.g. "normal",

good, fine, curious). Interestingly, participants' affective states did not affect recall,

which does not support Eich's (1995) mood hypothesis theory.

The non-significance of the results may be due to a number of these reasons.

However, it would seem that there were problems with the sample population. Upper

division psychology students may have more expert knowledge and hence, may have

guessed the purpose of the study. This may actually affect the results of the late

condition in which participants were told to wait quietly while the experimenter is setting

up; students from upper division psychology courses may see through the ruse. These

students from upper division courses may also be better at engaging in the encoding

strategies since they may have already developed better study strategies. The post hoc

comparisons suggest that there were differences in the performance of introductory

students versus upper division students. There were differential performances for the

participants depending on whether they were introductory psychology or upper division

students as is evidenced by very different means. However, due to the low and unequal









number of participants in each cell, caution should be taken when interpreting these

results.

It could also be possible that there were outliers in the data set, as is evidenced by

high standard deviations in some conditions such as the early, mismatch, and deep

experimental conditions for introductory to psychology students (SD = 7.071). These

outliers may have biased the data. In future follow-up studies, these outliers should be

replaced.

Another potential explanation for the non-significant results is not having enough

statistical power to detect the effects. There were eight participants in each of the

experimental cells which may not be enough participants to detect context change effects.

In spite of this, there were no ceiling effects or flooring effects. Follow-up studies should

include more participants and be cautious of participant recruitment.

General Conclusion

Despite all these possibilities, it could be that initial encoding strategy greatly

influences context change effect. It would make sense to think that information that is

more meaningful is less forgotten. In the real-life evidence, we often find that we forget

what we initially set out to do when moving to a new environment, such is the case when

going to obtaining a drink while watching television or retrieving an office supply in the

midst of typing. However, perhaps more often, we do not forget that we are going to

work or class as soon as we leave our homes or that we are going to return an important

phone call in the midst of the day. Perhaps this is because this information is meaningful;

we usually have a routine where we go to work or class at the same time every day or that

a phone call is very important. This information would be deeply encoded because it is

organized, important, and meaningful. The items or task that is usually forgotten due to









an environmental context change effect may be not as important as other activities that

are occurring; getting more paper for the printer or getting a drink in a midst of a

broadcast game may be not as important or of one's interest as working on the paper that

is due soon or the game that is playing on the television. We unconsciously prioritize our

tasks and information and hence, the items that are not as important are not processed as

much to prevent exhausting cognitive resources.

It is interesting to note that encoding strategies play a bigger role in incidental

memory, such as environmental context change, than was originally expected. In this

study, participants' encoding strategies were controlled; participants were required to

utilize a deep or shallow encoding strategy to study the words. It is important to

remember that without instruction, participants tend to overwhelmingly choose an initial

shallow encoding strategy (Delaney & Knowles, 2005; Sahakyan & Delaney, 2003;

Sahakyan, Delaney, & Kelley, 2004). This may not be adequate to explain the

discrepancies in the environmental context change literature, but future research should

examine the role of encoding strategies and the magnitude of environmental context

change effect.















LIST OF REFERENCES


Abernethy, E. M. (1940). The effects of changed environmental conditions upon the
results of college examinations. Journal ofPsychology, 10, 293-301.

Anderson, C. A. (1976). Coping behaviors as intervening mechanisms in the inverted-U
stress-performance relationship. Journal ofApplied Psychology, 61, 30-34.

Anderson, J. R., Bjork, R. A., & Bjork, E. L. (1994). Remembering can cause forgetting:
Retrieval dynamics in long-term memory. Journal ofExperimental Psychology:
Learning. Memory, & Cognition, 20, 1063-1087.

Baddeley, A. D. (1982). Domains of recollection. Psychological Review, 89, 708-729.

Benjamin, A. S., & Bjork, R. A. (2000). On the relationship between recognition speed
and accuracy for words rehearsed via rote versus elaborative rehearsal. Journal of
Experimental Psychology: Learnii.ng. Memory, & Cognition, 26, 638-648.

Bjork, R. A., & Richardson-Klavehn, A. (1988). On the puzzling relationship between
environmental context and human memory. In C. Izawa (Ed.) Current issues in
cognitive processes: The Tulane Flowerree Symposium on Cognition (pp. 313-
344). Hillsdale, NJ: Lawrence Erlbaum.

Bower, G. H., Clark, M. C., Lesgold, A. M., & Winzenz, D. (1969). Hierarchical retrieval
schemes in recall of categorized word lists. Journal of Verbal Learning and
Verbal Behavior, 8, 323-343.

Bradley, M. M., & Glenberg, A. M. (1983). Strengthening associations: Duration,
attention, or relations? Journal of Verbal Learning & Verbal Behavior, 22, 650-
666.

Cowan, N. (1988). Evolving conceptions of memory storage, selective attention, and their
mutual constraints with the human information-processing system. Psychological
Bulletin, 104, 163-191.










Craik, F. I., & Lockhart, R. S. (1972). Levels of processing: A framework for memory
research. Journal of Verbal Learning & Verbal Behavior, 11, 671-684.

Craik, F. I., & Tulving, E. (1975). Depth of processing and the retention of words in
episodic memory. Journal ofExperimental Psychology: General, 104, 268-294.

Dallett, K., & Wilcox, S. G. (1968). Contextual stimuli and proactive inhibition. Journal
ofExperimental Psychology, 78, 475-480.

Delaney, P. F., & Knowles, M. E. (2005). Encoding strategy changes and spacing effects
in the free recall of unmixed lists. Journal of Memory & Language, 52, 120-130.

Delaney, P. F., & Sahakyan, L. (2004). Unexpected costs of working memory: The impact
of contextual change. Manuscript submitted for publication.

Duffy, E. (1962). Activation and behavior. New York: Wiley.

Dulsky, S. G. (1935). The effect of a change of background on recall and relearning.
Journal ofExperimental Psychology, 18, 725-740.

Earles, J. L., Smith, A. D., & Parks, D. C. (1994). Age differences in the effects of
facilitating and distracting context on recall. Aging & Cognition, 1, 141-151.

Eich, J. E. (1980). The cue-dependent nature of state-dependent retrieval. Memory &
Cognition, 8, 157-173.

Eich, (J.) E. (1985). Context, memory, and integrated item/context imagery. Journal of
Experimental Psychology: Learniung. Memory, & Cognition, 11, 764-770.

Eich, (J.) E. (1995). Mood as a mediator of place-dependent memory. Journal of
Experimental Psychology: General, 124, 293-308.

Eich, J. E., Weingartner, H., & Stillman, R. C. (1975). State-dependent accessibility of
retrieval cues in the retention of a categorized list. Journal of Verbal Learning &
Verbal Behavior, 14, 408-417.

Einstein, G. O., & Hunt, R. R. (1980). Levels of processing and organization: Additive
effects of individual-item and relational processing. Journal ofExperimental
Psychology: Human Learning & Memory, 6, 588-598.

Fernandez, A., & Glenberg, A. M. (1985). Changing environmental contexts do not
reliably affect memory. Memory & Cognition, 13, 333-345.

Glenberg, A. M. (1997). What memory is for. Behavioral & Brain Sciences, 20, 1-55.









Glenberg, A. M., Schroeder, J. L., & Robertson, D. A. (1998). Averting the gaze
disengages the environment and facilitates remembering. Memory & Cognition,
26, 651-658.

Godden, D. R., & Baddeley, A. D. (1975). Context-dependent memory in two natural
environments: Land and underwater. British Journal ofPsychology, 66, 325-331.

Godden, D. R., & Baddeley, A. D. (1980). When does context influence recognition
memory? British Journal ofPsychology, 71, 99-104.

Groves, P. M., & Thompson, R. F. (1970). Habituation: A dual-process theory.
Psychological Review, 77, 419-450.

Isarida, T. (2005). Study time effect on free recall within and out of context. Memory, 13,
785-795.

Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression. Thousand
Oaks, CA: Sage.

Kane, M. J., Bleckley, M. K., Conway, A. R. A., & Engle, R. W. (2001) A controlled-
attention view of working memory capacity. Journal ofExperimental Psychology:
General, 130, 169-183.

Kane, M. J., & Engle, R. W. (2000). Working-memory capacity, proactive interference,
and divided attention: Limits on long-term memory retrieval. Journal of
Experimental Psychology: Learning. Memory, & Cognition, 26, 336-358.

Krafka, C., & Penrod, S. (1985). Reinstatement of context in a field experiment on
eyewitness identification. Journal ofPersonality & Social Psychology, 49, 58-69.

Mandler, G. (1975). Mind and emotion. New York: Wiley.

Masson, M. E., & McDaniel, M. A. (1981). The role of organizational processes in long-
term retention. Journal ofExperimental Psychology: Human Learning and
Memory, 7, 100-110.

Murnane, P., & Phelps, M. P. (1995). Effects of changes in relative cue size strength on
context-dependent recognition. Journal ofExperimental Psychology: Learninig.
Memory, & Cognition, 21, 158-172.

Nixon, S., & Kanak, N. J. (1981). The interactive effects of instructional set and
environmental context changes on the serial position effect. Bulletin of the
Psychonomic Society, 23, 139-142.









Park, D. C., Smith, A. D., Morrell, R. W., Puglisi, J. T., & Dudley, W. N. (1990). Effects
of contextual integration on recall of pictures in older adults. Journal of
Gerontology, 45, 52-57.

Perfect, T. J., & Dasgupta, Z. R. R. (1997). What underlies the deficit in reported
recollective experience in old age?. Memory & Cognition, 25, 849-858.

Sahakyan, L. (2004). The destructive effects of the "forget" instructions. Psychonomic
Bulletin & Review, 11, 555-559.

Sahakyan, L., & Delaney, P. F. (2003). Can encoding differences explain the benefits of
directed forgetting in the list-method paradigm?. Journal of Memory & Language,
48, 195-201.

Sahakyan, L., Delaney, P. F., & Kelley, C. M. (2004). Self-evaluation as a moderating
factor of strategy change in directed forgetting benefits. Psychonomic Bulletin &
Review, 11, 131-136.

Sahakyan, L., & Kelley, C. M. (2002). A contextual change account of the directed
forgetting effect. Journal ofExperimental Psychology: Learning. Memory, &
Cognition, 28, 1064-1072.

Saufley, W. H., Jr., Otaka, S. R., & Bavaresco, J. L. (1985). Context effects: Classroom
tests and contextual independence. Memory & Cognition, 13, 522-528.

Smith, A. D., Park, D. C., Earles, J. L. K., Shaw, R .J., & Whiting, W. L. (1998). Age
differences in context integration in memory. Psychology & Aging. 1, 21-28.

Smith, S. M. (1979). Remembering in and out of context. Journal ofExperimental
Psychology: Human Learning & Memory, 5, 460-471.

Smith, S. M. (1982). Enhancement of recall using multiple environmental contexts during
learning. Memory & Cognition, 10, 405-412.

Smith, S. M. (1984). A comparison of two techniques for reducing context-dependent
forgetting. Memory & Cognition, 12, 477-482.

Smith, S. M. (1985a). Effects of number of study environments and learning instructions
on free-recall clustering and accuracy. Bulletin ofPsychonomic Society, 23, 440-
442.

Smith, S. M. (1985b). Environmental context memory and recognition reconsidered.
Bulletin ofPsychonomic Society, 23, 173-176.

Smith, S. M. (1986). Environmental context-dependent recognition memory using a
short-term memory task for input. Memory & Cognition, 14, 347-354.










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). Oxford,
England: John Wiley & Sons.

Smith, S. M. (1995). Getting into and out of mental ruts: A theory of fixation, incubation,
and insight. In R.J. Sternberg & J.E. Davidson (Eds.), The nature of insight (pp.
229-251). Cambridge, MA, US: The MIT Press.

Smith, S. M., Glenberg, A., & Bjork, R. A. (1978). Environmental context and 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.

Smith, S. M., Vela, E., & Williamson, J. E. (1988). Shallow input processing does not
induce environmental context-dependent recognition. Bulletin ofPsychonomic
Society, 26, 537-540.

Sokolov, E. N. (1975). The neuronal mechanisms of the orienting reflex. In E.N. Sokolov
& O.S. Vinogradova (Eds.), Neuronal mechanisms of the orienting reflex (pp.
217-235). Hillsdale, NJ: Lawrence Erlbaum.

Strand, B. Z. (1970). Change of context and retroactive inhibition. Journal of Verbal
Learning & Verbal Behavior, 9, 202-206.

Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in
episodic memory. Psychological Review, 80, 359-380.

Turner, M. L., & Engle, R. W. (1986). Working memory. Proceedings of the Human
Factors Society, 30, 1273-1277.

Turner, M. L., & Engle, R. W. (1989). Is working-memory capacity task dependent?
Journal of Memory and Language, 28, 127-154.

Watkins, M. J. (1978). Engrams as cuegrams and forgetting as cue-overload: A cueing
approach to the structure of memory. In C.R. Puff (Ed.) The structure of memory
(pp. 347-372). New York: Academic Press.

Watkins, O. C., & Watkins, M. J. (1975). Build-up of proactive inhibition as a cue-
overload effect. Journal ofExperimental Psychology: Human Learning &
Memory, 1, 442-452.

Weingartner, H., Adefris, W., & Eich, J. E. (1976). Journal ofExperimental Psychology:
Human Learning & Memory, 2, 83-87.






46


Weiss, W., & Margolius, G. (1954). The effect of context stimuli on learning and
retention. Journal ofExperimental Psychology, 48, 318-322.

Wickens, D. D. (1970). Encoding categories of words: An empirical approach to
meaning. Psychological Review, 77, 1-15.

Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of stimulus to rapidity of
habit-formation. Journal of Comparative Neurology and Psychology, 18, 459-
482.















BIOGRAPHICAL SKETCH

I was born in Thailand in 1982; at that time, my parents and brother were on their

way to America. They were the "boat people" of Vietnam, refugees of the Vietnam War

in search of a better home and a hopeful future. Three months later, we eventually ended

up in Oklahoma City, Oklahoma, which I would come to regard as my home city and

would graduate from Northwest Classen High School in 2000. In 2003, I graduated from

Oklahoma State University with a Bachelor of Science degree in psychology and a minor

in biology. I entered the cognitive psychology program at the University of Florida in

2003.