|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
INFLUENCE OF TIME SPENT
IN AN ENVIRONMENT AND ENCODING STRATEGIES
ON THE ENVIRONMENTAL CONTEXT CHANGE EFFECT
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
Khanh N. Nghiem
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
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
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
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
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
Khanh N. Nghiem
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.
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,
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-
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
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
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.
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
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.
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,
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
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.
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
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.
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 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.
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).
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.
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.
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
A A A
0 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.
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
Time Mean S.D. Mean S.D.
Match .306 .073 .519 .189
Mismatch .343 .018 .494 .201
Match .344 .132 .600 .136
Mismatch .300 .046 .431 .194
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
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).
R Sq Linear = 0.316
R Sq Linear = 0.03
I I I I I
10 20 30 40 50
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
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
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 0 00
0 0 0 0
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
Time Mean S.D. Mean S.D.
Match .250 .050 .388 .085
Mismatch .342 .020 .450 .354
Match .276 .035 .608 .139
Mismatch .300 .000 .438 .144
. Tablel-3. Recall Mean Proportions and Standard Deviations for Upper-Division
Psychology Course Students
Time Mean S.D. Mean S.D.
Match .340 .065 .650 .173
Mismatch .350 .000 .508 .174
Match .367 .147 .625 .160
Mismatch .300 .071 .475 .203
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).
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.
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.
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
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
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.
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
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-
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,
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,
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,
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-
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.
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-
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