ArticlePDF Available

Is Expanded Retrieval Practice a Superior Form of Spaced Retrieval? A Critical Review of the Extant Literature


Abstract and Figures

onsider the following scenario: You are in the age range of most of the contributors to this volume and are at a neighborhood party. You are introduced to a person named Mark Finglestein. What would be the best procedure to learn this person's name so you are not embarrassed in future chance encounters in the neighborhood? One procedure would be to simply rehearse the name over and over again via massed practice. Of course, as students of learning and memory, we all know that this procedure is doomed to failure. Another procedure would be to space one's retrieval such that after every minute or so one attempts to retrieve the name Mark Finglestein. Because of the well-known benefits of spaced practice, this procedure is much more likely to succeed. A third procedure would be to gradually expand the intervals between the retrieval attempts. For example, you may first retrieve the name after 15 seconds, then 45 seconds, and then 2 minutes. This procedure takes advantage of the benefits of spacing but also maintains relatively high levels of retrieval success. There is evidence suggesting that this procedure may indeed be better than the simple spaced retrieval. In fact, the benefits of expanded retrieval have been at the center of considerable work in both educational and clinical settings. The present chapter reviews the evidence concerning the benefits of expanded retrieval over equal interval conditions. We will first provide some historical background on the spacing effect, since expanded retrieval may be considered a special case of spacing. We will then critically evaluate the evidence concerning expanded and equal interval retrieval practice. We will conclude with a discussion of some potential limits of expanded retrieval, theoretical implications, and possible avenues for future research.
Content may be subject to copyright.
Is Expanded Retrieval Practice a
Superior Form of Spaced Retrieval?
A Critical Review of the
Extant Literature
Consider the following scenario: You are in the age range of most of the
contributors to this volume and are at a neighborhood party. You are
introduced to a person named Mark Finglestein. What would be the best
procedure to learn this person’s name so you are not embarrassed in future chance
encounters in the neighborhood? One procedure would be to simply rehearse the
name over and over again via massed practice. Of course, as students of learning
and memory, we all know that this procedure is doomed to failure. Another
procedure would be to space one’s retrieval such that after every minute or so
one attempts to retrieve the name Mark Finglestein. Because of the well-known
benefits of spaced practice, this procedure is much more likely to succeed. A third
procedure would be to gradually expand the intervals between the retrieval
attempts. For example, you may first retrieve the name after 15 seconds, then
45 seconds, and then 2 minutes. This procedure takes advantage of the benefits
of spacing but also maintains relatively high levels of retrieval success. There is
evidence suggesting that this procedure may indeed be better than the simple
spaced retrieval. In fact, the benefits of expanded retrieval have been at the center
of considerable work in both educational and clinical settings.
The present chapter reviews the evidence concerning the benefits of expanded
retrieval over equal interval conditions. We will first provide some historical
background on the spacing effect, since expanded retrieval may be considered a
special case of spacing. We will then critically evaluate the evidence concerning
expanded and equal interval retrieval practice. We will conclude with a discussion
of some potential limits of expanded retrieval, theoretical implications, and
possible avenues for future research.
The spacing effect is one of the most ubiquitous findings in learning and memory.
Performance on a variety of tasks is better when the repetition of the to-be-learned
information is distributed as opposed to massed in presentation. This observation
was first formalized in Jost’s law, which states that “if two associations are of equal
strength but of different age, a new repetition has a greater value for the older one”
(McGeogh, 1943). Spacing effects occur across domains (e.g., learning perceptual
motor tasks vs. learning lists of words), across species (e.g., rats, pigeons, and
humans), across age groups and individuals with different memory impairments,
and across retention intervals of seconds to months (see Cepeda, Pashler, Vul,
Wixted, & Rohrer, 2006; Crowder, 1976; Dempster, 1996, for reviews). In this
light, it is interesting that spacing effects have not received much attention in
Cognitive Psychology textbooks. In fact, in our sampling of seven such textbooks,
only one had a section dedicated to this topic, while virtually all cognitive text-
books discussed mnemonic techniques such as the pegword or method of loci.
Given the power and simplicity of implementing spaced practice, we clearly hope
this changes in the future.
As a sidebar, we felt it was particularly fitting to discuss the spacing effect as
part of a book honoring the contributions of Roddy Roediger. Roddy’s mentor,
Robert Crowder (1976) devoted considerable discussion (over 40 pages) of his
classic book, The Principles of Learning and Memory, to the effects of spacing
on learning and memory. Moreover, Arthur Melton, Roddy’s academic grand-
father, wrote a definitive piece on this topic in 1970. As noted in Melton’s paper,
spacing effects, like many things, can be traced back to observations by William
James (Roddy’s great-great-great grandfather) concerning the benefits of alter-
nating swimming in the summer and skating in the winter. In addition to his
academic lineage, spacing effects are also quite consistent with Roddy’s functional-
ist approach in his work and the importance of making a connection between
practical implications and basic experimental research. Recently, Roddy has been
exploring the application of cognitive principles to educational practice, and has
already made important contributions in this domain (see, for example, Roediger
& Karpicke, 2006). Expanded retrieval and the spacing effect have been at the
center of considerable work in educational settings, and Roddy has been exploring
this issue in a number of recent projects (e.g., Karpicke & Roediger, 2005).
Although spaced presentation is clearly beneficial in most situations, as with every-
thing, there are some limitations. For example, there is evidence of a spacing
by retention interval interaction. Specifically, massed items are actually better
retained after a short retention interval, whereas spaced items are better retained
after a long retention interval. This observation was first obtained by Peterson,
Wampler, Kirkpatrick, and Saltzman (1963), and was replicated by Glenberg (1977)
and Balota, Duchek, and Paullin (1989).
There are numerous theoretical accounts of the spacing effect (see Cepeda
et al., 2006, for a review); however, the spacing by retention interval interaction
would appear to be most consistent with an encoding variability account. Because
of the relevance of this theoretical framework to our discussion of the benefits
of expanded retrieval in a later section, we will briefly describe the encoding
variability account of the spacing by retention interval interaction. According to
encoding variability theory, performance on a memory test is dependent upon the
overlap between the contextual information available at the time of test and the
contextual information available during encoding. During massed study, there is
relatively little time for contextual elements to fluctuate between presentations
and so this condition produces the highest performance in an immediate memory
test, when the test context strongly overlaps with the same contextual information
encoded during both of the massed presentations. In contrast, when there is
spacing between the items, there is time for fluctuation to take place between the
presentations during study, and hence there is an increased likelihood of having
multiple unique contexts encoded. Because a delayed test will also allow fluctu-
ation of context, it is better to have multiple unique contexts encoded, as in the
spaced presentation format, as opposed to a single encoded context, as in the
massed presentation format.
Because of evidence regarding a contextual encoding deficit in older adults
(see Burke & Light, 1981; Duchek, 1984), Balota et al. (1989) investigated the
spacing by retention interval interaction in young and older adults, and found that
both groups produced clear evidence of the interaction, with older adults being
overall lower. More importantly, as suggested in Crowder (1976), Balota et al. fit
their data to Estes’ (1955) stimulus sampling model (a precursor of encoding
variability theory), which was developed to account for aspects of extinction and
spontaneous recovery, among other findings, in the animal learning domain.
Interestingly, when animals are placed in an environment at varying times after
extinction of a response, the response appears to spontaneously recover, even
though it had been extinguished earlier. The results of this simple modeling
endeavor indicated that older adults were different from younger adults in two
parameter values. Specifically, older adults were less likely to store contextual
elements in memory and elements also fluctuated between available and unavail-
able states across time at a slower rate in older adults than younger adults. As
shown in Figure 6.1, changes in these two parameters of the Estes’ model nicely fit
the Balota et al. data (also see Spieler & Balota, 1996, for an extension to an
implicit memory task). Thus, the spacing by retention interval interaction nicely
extends to individuals with different levels of memory competence, and converges
on a particular theoretical account of the spacing effect that emphasizes contextual
drift (encoding variability) across time.
The spacing by retention interval interaction indicates that there are limita-
tions to the spacing effect. This tradeoff between lag (the amount of spacing
between two events) and retention interval has led some researchers (e.g., Greene,
1992) to speculate that the ratio between spacing and retention interval is the
critical variable. Although there is some tradeoff between retention interval and
spacing, there are probably some limits even here. For example, if one considers
a spaced interval of 1 year between two repetitions, one may not expect much
benefit even at a very long retention interval, since the initial trace may have either
decayed or been completely overwhelmed by interfering material. Of course, this
FIGURE 6.1 Percent correct recall as a function of group and retention interval from
Balota et al. (1989), along with the predictions from Estes’ (1955) stimulus fluctuation
leads us to an important question. Are there ways to maintain the strength of the
initial encoding so as to minimize the disruptive effects at long lags? Enter
expanded retrieval practice. The goal in this procedure is to gradually increase the
lag between retrieval events, thereby maintaining relatively high levels of retrieval
success at long lags.
In an often-cited chapter, Landauer and Bjork (1978) were the first to carefully
explore expanded retrieval in two well-controlled experiments. In their first
experiment, subjects were presented a deck of cards that included fictitious first
and second names of individuals during the acquisition phase. For present pur-
poses, we will focus on the trials in which subjects were first presented both
the first and second names intact followed at varying schedules of receiving cards
with only the first name of the study pair as a retrieval cue for the second name.
For example, in an expanded retrieval schedule, subjects may receive the first
name as a retrieval cue at 1, 4, and 10 intervening cards versus an equal interval
condition of 5, 5, and 5 intervening cards. A massed condition (study of the
name followed by three immediate tests) was used to obtain a baseline estimate
of the influence of simple repetition without spacing. The results of a later
cued recall test yielded a large benefit for both the expanded and equal interval
condition over the massed condition. More importantly, there was a small, but
significant effect (approximately a 5% benefit, based on interpolation from their
Figure 2) of the expanded condition over the equal interval condition in the final
cued recall test.
The Landauer and Bjork (1978) results have been quite influential and would
appear to naturally maximize the benefits of spacing. They originally suggested
that the benefit of expanded retrieval involves insuring successful retrieval at longer
and longer intervals, thereby increasing the difficulty of retrieval, while still yield-
ing success. One could also view these benefits within an encoding variability
framework described above. Specifically, with the passage of time and intervening
information, there is an increased likelihood of greater fluctuation of context,
and hence a greater likelihood of distinct encodings. The notion here is that
in a long-term retention test, stimuli that have more distinct encodings, afforded
by the longest interval in the expanded condition, would produce a greater
benefit compared to stimuli with less distinct encodings due to shorter retention
Both of the above accounts emphasize the additional benefit beyond success-
ful retrieval in the expanded condition compared to the equal interval condition.
In particular, the reason there is an advantage in the expanded condition is that the
participants are indeed retrieving items in the face of a longer retention interval
during the final retrieval attempt in the expanded compared to the equal interval
condition. Importantly, Landauer and Bjork (1978) provided retrieval success data
during initial list presentation to directly examine this. Based on interpolation
from their Figure 1, Table 6.1 displays performance of an expanded condition
compared to an equal interval condition estimated from Landauer and Bjork’s
first experiment. As shown, there is an initial benefit of expanded retrieval that
actually decreases across subsequent retrieval attempts. Hence, the benefit of
expanded over equal interval is larger during the acquisition phase than in final
recall. This is somewhat counterintuitive because one might expect an advantage
of expanding over equal interval schedules above and beyond the retrieval success
advantage produced during the acquisition phase. Specifically, at the last retrieval
event, the retrieved items have persisted in the face of a longer retention interval
and more intervening information in the expanded than in the equal interval
condition. On the other hand, the “expanded retrieval effect” may be due to the
fact that the first retrieval success occurs soon after study, permitting greater
retrieval success in the expanding relative to the equal interval condition, as
opposed to greater retrieval efficiency at the longest delays. Of course, this could
ultimately overestimate the benefits of expansion. We will return to this possibility
The Landauer and Bjork (1978) results are clearly provocative, and the bene-
fits of expanded retrieval have been explored in both educational and clinical
settings. Below we will provide a review of the influence of expanded retrieval
in these areas. It will become clear that although expanded retrieval is clearly
a useful memory aid, there has been relatively little controlled work, akin to
Landauer and Bjork’s, that attempts to isolate the influence of expansion over
comparable equal spaced conditions. Given the potential importance of this
procedure, this is surprising.
Before turning to the review of the literature, we believe that it is noteworthy
to identify the conditions that are important to make strong inferences regarding
the specific role of expanding retrieval, many of which were available in Landuaer
and Bjork’s seminal study. For example, it is important to include at least three
levels of spacing: an expanded condition (e.g., 1, 4, 10), an equal interval condition
that matches the average spacing of the expanded condition (e.g., 5, 5, 5), and a
massed condition (e.g., 0, 0, 0). With these three levels, one can both measure the
influence of expansion with spacing equated (e.g., 1, 4, 10 vs. 5, 5, 5), and also the
benefits of spaced (1, 4, 10 and 5, 5, 5) over massed practice (e.g., 0, 0, 0) to insure
one obtains a clear spacing effect. In addition, it is useful to measure retrieval
success during the initial acquisition phase to obtain an estimate of the original
level of learning. As we shall discuss below, it is preferable to include two different
delays for the final memory test, because it appears that one can obtain different
effects depending upon the delay. Of course, this has some resemblance to the
spacing by retention interval interaction. Finally, it would also be useful to include
a study alone condition with the same spacing and testing conditions to measure
the specific influence of retrieval during study. Unfortunately, as noted, these
conditions have rarely been available in the same study to isolate the specific
benefits of expanded retrieval.
TABLE 6.1 Interpolated Results from Landauer and Bjork (1978)
Retrieval attempt Final recall
Expanded .61 .55 .50 .45
Equal interval .42 .42 .43 .40
The additional benefit of expanded retrieval over equal interval spacing obviously
has important implications for educational settings. Although Landauer and Bjork’s
(1978) work sparked considerable interest in applying expanded retrieval to edu-
cational settings, Roddy informed us of relevant work nearly 40 years before
Landauer and Bjork’s study. Specifically, Spitzer (1939) incorporated a form of
expanded retrieval in a study designed to assess the ability of sixth graders to learn
science facts. Impressively, Spitzer tested over 3600 students in Iowa—the entire
sixth-grade population of 91 elementary schools at the time. The students read two
articles, one on peanuts and the other on bamboo, and were given a 25-item
multiple choice test to assess their knowledge (such as “To which family of plants
does bamboo belong?”). Spitzer tested a total of nine groups, manipulating both
the timing of the test (administered immediately or after various delays) and the
number of identical tests students received (one to three). Spitzer did not
incorporate massed or equal interval retrieval conditions, but he had at least two
groups that were tested on an expanding schedule of retrieval, in which the inter-
vals between tests were separated by the passage of time (in days) rather than by
intervening to-be-learned information. For example, in one of the groups, the first
test was given immediately, the second test was given seven days after the first test,
and the third test was given 63 days after the second test. Thus, in essence, this
group was tested on a 0-7-63 day expanding retrieval schedule. Spitzer compared
performance of the expanded retrieval group to a group given a single test 63 days
after reading the original article. On the first (immediate) test, the expanded
retrieval group correctly answered 53% of the questions. After 63 days and two
previous tests, their score was still an impressive 43%. The single test group cor-
rectly answered only 25% of the original items after 63 days, giving the expanded
retrieval group an 18% retention advantage. This is quite impressive, given that
this large benefit remained after a 63-day retention interval. Similar beneficial
effects were found in a group tested on a 0-1-21 day expanded retrieval schedule
compared to a group given a single test after 21 days. Of course, this study does not
decouple the effects of testing from spacing or expansion, but the results do clearly
indicate considerable learning and retention using the expanded repeated testing
procedure. Spitzer concluded that “. . . examinations are learning devices and
should not be considered only as tools for measuring achievement of pupils”
(p. 656, italics added), and we are sure Roddy would agree (see Roediger &
Karpicke, 2006).
Since Spitzer (1939), there was very little work was published on expanded
retrieval using educationally-relevant materials for 45 years until Rea and
Modigliani (1985) tested the effectiveness of expanded retrieval in a third-grade
classroom setting. In separate conditions, students were given new multiplication
problems or spelling words to learn. The problem or word was presented audio-
visually once and then tested on either a massed retrieval schedule of 0-0-0-0 or an
expanding schedule of 0-1-2-4, in which the intervals involved being tested on old
items or learning new items. After each test trial for a given item, the item was
re-presented in its entirety so students received feedback on what they were
learning. Performance during the learning phase was at 100% for both spelling
words and multiplication facts. On an immediate final retention test, Rea and
Modigliani found a performance advantage for all items—math and spelling—
practiced on an expanding schedule compared to the massed retrieval schedule.
They suggested, as have others, that spacing combined with the high success rate
inherent in the expanded retrieval schedule produced better retention than
massed retrieval practice. However, as in Spitzer’s study, Rea and Modigliani did
not test an appropriate equal interval spacing condition. Hence, their finding that
expanded retrieval is superior to massed retrieval in third graders could simply
reflect the superiority of spaced versus massed rehearsal—in other words, the
spacing effect.
More recently, Cull (2000) compared expanded retrieval to equal interval
spaced retrieval in a series of four experiments designed to mimic typical teaching
or study strategies encountered by students. He examined the role of testing
versus simply restudying the material, feedback, and various retention intervals on
final test performance. Paired associates (an uncommon word paired with a com-
mon word, such as bairn–print) were presented in a manner similar to the flash-
card techniques students often use to learn vocabulary words. The intervals
between retrieval attempts of to-be-learned information ranged from minutes in
some experiments to days in others. Interestingly, across four experiments, Cull
did not find any evidence of an advantage of an expanded condition over a uniform
spaced condition (i.e., no significant expanded retrieval effect), although both
conditions consistently produced large advantages over massed presentations. He
concluded that distributed testing of any kind, expanded or equal interval, can be
an effective learning aid for teachers to provide for their students.
Karpicke (2004) investigated the effectiveness of using expanded retrieval to
learn material commonly encountered by students—prose passages. Participants
were given 7 minutes to read an encyclopedia article and practice free recalling
the article on an expanding or equal interval schedule. An important feature of
Karpicke’s study was that he matched the timing of the first recall attempt in the
acquisition phase in the expanded and equal interval conditions. As noted, in
most expanded retrieval studies, the first practice for the expanded items comes
sooner than practice for the equal interval items. Hence, in Karpicke’s study, the
first recall attempt for both equal and expanded conditions was either immediate
(given just after reading the article) or delayed (given after 7 minutes of a
distractor task). He tested a total of four recall schedules in the acquisition
phase: 0-2-4, 0-2-2, 2-2-4, 2-2-2. The final free recall test was given 1 week later.
In both the immediate and delayed practice conditions, Karpicke found no
difference in memory performance when expanded and equal interval schedules
were matched on the timing of the first retrieval attempt. This was true in the
acquisition phase as well as the final free recall test. However, one might com-
pare the 0-2-4 condition to the 2-2-2 conditions to investigate the expanded
retrieval effect.
Interestingly, Karpicke (2004) did find an advantage for the expanded condi-
tion in both the acquisition phase as well as the final free recall test, similar to
Landauer and Bjork (1978). Because there was no delay in the first retrieval event
in the immediate retrieval conditions in this study, and the 0-2-4 and 0-2-2 pro-
duced equivalent and better performance than the 2-2-2 and the 2-2-4, Karpicke
points out that it is difficult to isolate the influence of expansion from the immedi-
acy of the first recall event. In this light, it would have been useful to have 0-1-2-3
and 0-2-2-2 conditions for comparison. In any case, Karpicke’s findings support
the notion that any expanded retrieval effect at a final test is strongly linked to
greater retrieval success for the expanded condition compared to the equal inter-
val condition during the first acquisition phase. When these two conditions were
matched in performance during the acquisition phase, the expanded retrieval
effect did not emerge.
Researchers have also attempted to extend the utility of expanded retrieval to
other “real world” settings such as in our original example of learning the name
Mark Finglestein (remember Mark?) at a neighborhood party. Morris and Fritz
(2000; see also Morris & Fritz, 2002; Morris, Fritz, & Buck, 2004) modified the
expanded retrieval technique to create a “name game” to help individuals learn the
names of members in their group. In the name game, an individual introduces
herself, while a group leader writes their name on a board and then erases it once
the other group members had read it. A second individual then repeats the first
person’s name aloud and introduces himself to the group. The third individual
then repeats the names of the first and second individuals (in that order) and adds
her name, etc. This technique of repeating previous names (starting with the first
person’s name) then adding a new name to the group is followed until all group
members are introduced. Thus, the interval between hearing a given name pro-
gressively expands as each new person is introduced. In the Morris and Fritz (2000)
study, participants were given a final recall test on the names after 30 minutes and
again after 2 weeks and 11 months. Performance in the name game condition was
compared to a condition in which individuals were simply paired with another
group member and then asked to introduce their partners aloud to the group. In
two experiments, participants recalled significantly more full names from the name
game condition than the partner condition, even after 11 months. This condition
also produced superior performance compared to when the group leader simply
read the previous names back after each new introduction without ever asking
individuals to retrieve the names. In a follow-up study, Morris and Fritz (2002)
found that a reversed version of the name game, in which names were recalled
in reverse chronological order, produced better performance than the original
version used in the previous study. They noted that the reversed name game
more closely mimics the technique originally employed by Landauer and Bjork
(1978), and it was more effective than the original name game because there was
less chance of forgetting a new name when the delay between learning and
retrieval was reduced. Although the name game clearly does work, there is again
no comparable equal interval condition to examine if the schedule is the critical
dimension or the spacing.
The studies reviewed above specifically cited a desire to apply the expanded
retrieval technique to nonlaboratory settings and explored its applications to a
variety of learning situations (science and multiplication facts, spelling words, and
first and last names). The studies found an advantage for expanded retrieval com-
pared to information that is simply repeated or reread (Cull, 2000; Morris & Fritz,
2000, 2002), tested only once (Spitzer, 1939), or tested on a massed retrieval
schedule (Cull, 2000; Rea & Modigliani, 1985). In addition, Karpicke (2004) pro-
vides intriguing evidence that provides support for the utility of expansion for text
materials, but overall his results primarily show the importance of the immediacy
of a first testing session. Only Cull (2000) directly compared the utility of expanded
retrieval to other schedules of spaced retrieval and this study did not obtain any
benefits of expanded retrieval over an equal interval schedule. In a later section,
we will further discuss the Cull study and other studies that have attempted to
decouple the influence of spacing from expansion in more traditional laboratory
Another area where expanded retrieval has been viewed as having considerable
potential is as a cognitive rehabilitation technique for various memory-impaired
populations. This area has been very active. The specific appeal of expanded
retrieval is that the benefits of expanded retrieval are relatively automatic and
nonstrategic, as opposed to teaching memory impaired individuals mnemonic
techniques such as the pegword method or method of loci, which demand
considerable strategic/attentional control processes.
Schacter, Rich, and Stampp (1985) utilized the expanded retrieval procedure
to improve memory performance in four patients with cognitive impairment due
to a variety of etiologies. In this study, subjects first studied faces and associated
characteristics (i.e., names, hometowns, occupations, hobbies) and then utilizing
the face cue, engaged in retrieval practice for the associated characteristics at
expanded time intervals. Initially subjects were given a verbal prompt to engage
in retrieval practice and over time the verbal prompt was removed to see if sub-
jects would engage in expanded retrieval practice on their own. The results indi-
cated that memory performance for the associated characteristics improved after
expanded retrieval practice relative to baseline performance as measured prior
to the introduction of the expanded retrieval strategy. Importantly, two out
of the four subjects were relatively successful in spontaneously using expanded
retrieval practice without prompts. Thus, Schacter et al. concluded that expan-
ded retrieval practice might be a particularly effective strategy for memory
enhancement in cognitively impaired populations.
The vast majority of the work employing the use of expanded retrieval in
clinical populations has been conduced by Camp and his colleagues, with indi-
viduals in various stages of Alzheimer’s disease (AD; for a review, see Camp, Bird,
& Cherry, 2000). Given that memory loss is the hallmark symptom of Alzheimer’s
disease, targeting this population with such a behavioral intervention could prove
quite beneficial to both patients and caregivers. In an effort to apply Landauer
and Bjork’s (1978) expanded retrieval practice to the AD population, Camp and
colleagues adopted a clinical protocol for teaching AD patients new information
using spaced retrieval practice (Camp, 1989; Camp & McKitrick, 1992). In this
procedure, subjects are first given one piece of information to remember (e.g., the
name of a staff person) and are then tested for immediate recall. If this immediate
recall attempt is successful, then the next retrieval attempt is queried after 5 s, and
then expanded to 10 s, 20 s, 40 s, 60 s. After successful retention of the item for
60 s, intervals are increased by 30 s (90 s, 120 s, etc.). If a person does not
successfully retrieve the information, he/she is given the correct response, asked to
immediately repeat the correct response, and then is tested at the last successful
interval. Expanded retrieval attempts are again initiated from that point forward
on successful retrieval events.
As Camp, Foss, Stevens, and O’Hanlon (1996) describe, there are some key
features of this spaced retrieval strategy that promote success in memory-impaired
patients. First, the expansion of retrieval attempts occurs over time, not over
intervening to-be-remembered items. Thus, the time intervals are typically filled
with social conversation or other non-related activities. Hence, as Camp and
Mattern (1999) argue, the expanded retrieval training can be efficiently used
within a therapy session wherein the intervening time intervals can be used for
other therapeutic activities. Second, this technique is analogous to a shaping pro-
cedure wherein an association is formed between a stimulus (i.e., the retrieval
query) and a single response (i.e., the name). Because the learning occurs very
gradually over time, the initial retrieval attempts are likely to be successful, even
for individuals in the moderate stages of AD. Third, the expanded retrieval strat-
egy incorporates the neurorehabilitation technique of errorless learning (Wilson,
Baddeley, Evans, & Shiel, 1994). Wilson and colleagues have argued that memory-
impaired populations show better long-term retention for information when they
are not allowed to make errors during training (Baddeley & Wilson, 1994; Wilson
& Evans, 1996). The notion is that explicit memory is necessary for error recogni-
tion and elimination during learning. When the explicit system is deficient, then
errors may be implicitly incorporated into the learning and thus each repetition of
an erroneous response may serve to further strengthen that incorrect response,
thereby hindering the retention of the correct information. Thus, the expanded
retrieval procedure used by Camp and colleagues requires that when there is
a failure of retrieval, the correct response be given immediately and repeated by
the subject. Finally, the implementation of the expanded retrieval strategy
requires little cognitive effort and/or resources on the part of the learner. Camp
and colleagues have argued that expanded retrieval training makes use of implicit
memory, which is relatively spared even in the later stages of AD (e.g., Balota &
Ferraro, 1996; Camp et al., 2000; Faust, Balota, & Spieler, 2001). Thus, it is ideally
suited for memory-impaired patients.
Indeed there have been numerous reports in the literature of the successful
use of expanded retrieval practice in teaching new information to individuals in
the relatively advanced stages of AD. Specifically, Camp and colleagues have used
this technique to teach AD individuals various types of information, such as names
of common objects (e.g., McKitrick & Camp, 1993), face-name associations (e.g.,
Camp & Schaller, 1989), object-location associations (e.g., Camp & Stevens, 1990),
and even prospectively remembering to perform a task (e.g., McKitrick, Camp, &
Black, 1992). Moreover, long-term retention of information has been demonstrated
over several days in some cases (e.g., Camp et al., 1996). For example, in the latter
study, Camp et al. employed an expanding retrieval strategy to train 23 individuals
with mild to moderate AD to refer to a daily calendar as a cue to remember to
perform various personal activities (e.g., take medication). Following a baseline
phase to determine whether subjects would spontaneously use the calendar,
spaced retrieval training was implemented by repeatedly asking the subject the
question, “How are you going to remember what to do each day?” at expanding
time intervals. The results indicated that 20/23 subjects did learn the strategy
(i.e., to look at the calendar) and retained it over a 1-week period.
There have been numerous other studies utilizing this protocol of expanded
retrieval practice to induce memory-impaired patients to learn clinically relevant
behaviors. For example, this technique has been employed to teach a demented
client with dysphagia (i.e., a swallowing disorder) to use a compensatory strategy of
alternating bites of food with sips of liquid to prevent aspiration (Brush & Camp,
1998a). Expanded retrieval training has been applied to attaining goals during
speech therapy sessions (Brush & Camp, 1998b) with demonstrated retention over
a 4-week period in demented clients. Patients with dementia associated with
Parkinson’s disease have successfully learned new motor tasks (Hayden & Camp,
1995) and patients with dementia associated with HIV have been trained in the
use of external aids (Lee & Camp, 2001) with expanded retrieval practice.
Likewise, a client with AD was able to learn the names of 11 members of his social
group with a combined intervention of errorless learning and expanded retrieval
(Clare et al., 2000). There has even been a case report of an individual in the early
stages of AD who trained himself to spontaneously use expanded retrieval practice
to remember new information (Riley, 1992). Thus, the benefits of expanded
retrieval have been widely documented in the literature across various targeted
behaviors and clinical populations.
What is somewhat surprising in reviewing all of these studies is that there has
been no attempt to systematically compare expanded retrieval practice with other
schedules of spaced retrieval practice, such as equal interval spacing or even
massed spacing of retrieval practice. In most of these studies, performance after
expanded retrieval practice is simply compared relative to baseline performance
(e.g., Camp et al., 1996; Cherry, Simmons, & Camp, 1999). More recently,
Bourgeois et al. (2003) compared expanded retrieval training with a modified
cued hierarchy training in teaching demented individuals to use various external
memory aids. The modified cued hierarchy training is initiated in a similar way
as the expanded retrieval training in that patients are first given the target informa-
tion (e.g., when you want to know what activity to perform today—look at your
activity list) and then immediately queried for that information. If patients cannot
immediately give the correct response or give an incorrect response, they are given
a hierarchy of cues in the following order until the correct response is given:
semantic (“Something to look at”), phonemic (the first syllable of “activity list”),
visual (point to list), tactile (touch list), imitation (“I look at my activity list”). No
cues are given in the expanded retrieval training. The results indicated that both
strategies improved usage of external aids, but the expanded retrieval strategy was
more effective. Of course, the latter study simply indicates that spaced retrieval
practice leads to better retention than providing a hierarchy of cues. It did not
address the question of which specific pattern of spacing leads to the greatest
benefit in long-term retention.
This issue of isolating which component of expanded retrieval is critical with a
clinical sample was more systematically addressed in a recent study by Hochhalter,
Overmier, Gasper, Bakke, and Holub (2005). Individuals with AD were presented
pill names (Exp. 1; n = 10) under five different schedules of retrieval practice:
massed, uniform distributed (i.e., equal interval), spaced/expanded adjusted based
on performance (similar to the Camp protocol), expanded without adjustment,
and random. The results indicated that only 6/10 individuals showed long-term
retention for at least one pill name and there was no difference in the number of
“learners” across the different schedules of retrieval practice. Thus the spaced/
expanded retrieval condition did not show a benefit relative to any of the other
schedules of practice. If anything, more subjects showed long-term retention in
the random condition (n = 5) than the spaced/expanded condition (n = 3) and also
had fewer errors during training.
The Hochhalter et al. (2005) study is clearly informative, but the relatively
small number of subjects and more variable applied therapeutic setting makes
it difficult to draw any firm conclusions. However, as described below, Balota,
Duchek, Sergent-Marshall, and Roediger (2005) provide a study comparing equal
interval and expanded practice in a laboratory context with a large set of healthy
young, older adults, and individuals with early stage Alzheimer’s disease and come
to a very similar conclusion as the Hochhalter et al. study regarding comparisons
of equal interval and expanded retrieval.
It is interesting to note that in their initial study, Schacter et al. (1985) acknow-
ledged that it was unclear whether it was the actual expanding pattern of retrieval
that led to the performance benefit in their study or whether it was merely due to
the simple repetition of retrieval attempts. They argued that to some extent the
latter question is not really clinically important given that expanded retrieval prac-
tice does indeed work for memory-impaired patients. Clearly, one would agree
with this statement on an individual patient basis, where the goal of treatment is to
enhance memory for important personal information. It is evident from the litera-
ture that having memory-impaired individuals engage in an expanded retrieval
strategy does improve memory performance under a variety of situations, at least
relative to baseline (no treatment) performance (Camp, Bird, & Cherry, 2000).
Furthermore, one can see how the gradual expansion of recall attempts with the
initial accompanying retrieval success would be relatively easy to implement in a
clinical setting or in a home setting by caregivers (McKitrick & Camp, 1993).
However, from a theoretical perspective, it is also critical to examine the proper-
ties of the most effective spacing of retrieval practice to enhance long-term reten-
tion. This research could also provide evidence for designing the best intervention
strategies for memory-impaired individuals.
As reviewed in the previous two sections, there is clear evidence that expanded
retrieval practice has considerable potential in both educational and applied set-
tings. The question that we will now turn to is an examination of how much of this
evidence is specific to “expanded” retrieval. This of course depends on what is the
appropriate baseline. As in Landauer and Bjork’s (1978) original paper, we believe
the appropriate baseline to be a comparably matched equal interval schedule. In
this way, one can directly investigate if there is an additional benefit of expansion
over mere spacing. For these studies, we are forced to return to the laboratory
where more control of critical variables is available.
In one of the most comprehensive studies in this area, Cull, Shaughnessy, and
Zechmeister (1996) explored the benefits of expanded retrieval over uniform spa-
cing in a series of five experiments. They compared the expanded schedules of
1-5-9, with the equal interval schedule of 5-5-5, and included a 0-0-0 massed
presentation condition, as a baseline to measure the benefit of spacing on a later
final recall test. As shown in Table 6.2, the results generally support the benefit of
expanded retrieval over equal interval conditions. Experiments 1 and 4 produced
reliable expanded retrieval benefits compared to an equal interval condition,
whereas the results from Experiments 2 and 3 were in the same direction, but did
not reach significance. Experiment 5 produced near ceiling effects in both spaced
conditions. Somewhat surprisingly, the comparison of the equal interval condition
to the massed condition did not consistently produce spacing effects. A test for the
spacing effect was not directly provided for Experiments 1 and 2, and in Experi-
ments 3 and 4, these two conditions produced identical performance. Only, their
fifth experiment provided strong evidence for a spacing effect, but this experiment
did not provide evidence for an expanded retrieval effect, because of ceiling
Cull (2000, Exp. 1) provided a direct follow-up study to the initial Cull et al.
TABLE 6.2 Results from Cull et al. (1996), and Cull (2000) Studies
Experiment Presentation schedule
1-5-9 5-5-5 0-0-0
Cull et al. (1996)
1. Test only .33 .23 .17
2. Test only .27 .18 .11
3. Test only .34 .28 .28
4. Test only .72 .59 .59
5. Test–study .91 .91 .62
Cull (2000, Exp. 1)
Test only .38 .34 .18
Study only .27 .29 .14
Test–study .48 .49 .19
(1996) study. He included the same three spacing schedules used in the ear-
lier Cull et al. study, along with three different types of study. In essence, his
first experiment included three different experiments, as these conditions were
between participant manipulations. In the study-only condition, the cue and
response item of a paired associate were both presented on each repetition for 8 s.
In the test-only condition, the cue for retrieval was presented on each occasion for
8 s. In the study–test condition, the cue was presented for 6 s for retrieval, followed
by the response item for 2 s. As shown in Table 6.2, Cull did not obtain an advan-
tage for the expanded over the equal interval condition in any of these conditions.
As noted earlier, Cull also explored more educationally relevant manipulations in
three additional experiments. None of these experiments afforded any benefit of
expanded retrieval over a comparable equal interval condition.
Carpenter and DeLosh (2005, Exp. 2) have recently investigated face-name
learning under massed, expanded (1-3-5), and equal interval (3-3-3) conditions.
This study also involved study and study and test procedures during the acquisition
phase. Carpenter and DeLosh found a large effect of spacing, but no evidence of a
benefit of expanded over equal interval practice. In fact, Carpenter and DeLosh
reported a reliable benefit of the equal interval condition over the expanded
retrieval condition. Although Carpenter and DeLosh did report a reliable (20%)
benefit of expanded retrieval during the acquisition phase, they did not break this
down by each retrieval event to determine the size of the difference at the last
retrieval event in the expanded condition.
In addition to studies with healthy young adults, there have been two recent
studies that have explored the benefits of expanded retrieval with healthy older
adults and early stage Alzheimer’s disease. Balota et al. (2005) compared a massed
presentation condition (0-0-0-0-0-0) with an equal interval (0-0-3-3-3) and an
expanded retrieval schedule (0-0-1-3-5). Because this study included healthy
young adults, older adults, and individuals with Alzheimer’s disease, who have
quite a wide range of memory ability, Balota et al. had subjects engage in two
massed retrieval attempts before each of the three schedules to insure all groups
were off the floor in their final recall performance. The results of performance
during the acquisition phase are shown in Figure 6.2. Here one can see that the
expanded retrieval condition indeed produces a benefit over the equal interval
condition that persists until the last retrieval event for all groups of participants.
Hence, because performance on the last retrieval event reflects longer mainten-
ance of the memory in the 0-0-1-3-5 condition than the 0-0-3-3-3 condition, one
would expect a benefit to persist in the final cued recall tests. This was not what the
results indicated. Specifically, as shown in Figure 6.3, all three groups produced
large spacing effects in a final cued recall test; however, there was no evidence of a
reliable difference between the expanded and equal interval conditions for any of
the groups. Moreover, the lack of a difference in final recall between the expanded
and equal interval condition was replicated in two subsequent experiments, in
which corrective feedback was given to participants. Hence, the Balota et al.
results converge nicely with the recent Cull (2000) and Carpenter and DeLosh
(2005) studies and extends this pattern to both healthy older adults and individuals
with early stage Alzheimer’s disease.
Of course, one might be concerned that the Balota et al. (2005) and Carpenter
and DeLosh (2005) study used the 3-3-3 and 1-3-5 comparison. Logan (2004)
conducted a comprehensive study that compared three schedules of retrieval
practice in healthy young and older adults, i.e., 0-0-0 versus 1-2-3 versus 2-2-2;
0-0-0 versus 1-3-5 versus 3-3-3; 0-0-0 versus 1-3-8 versus 4-4-4. An important
additional contribution of the Logan study is that she also included both an
FIGURE 6.2 Mean proportion correct during the acquisition phase for the Young (top
panel), Old (middle panel) and DAT (bottom panel) individuals, as a function of spacing,
and retrieval attempt from Balota et al. (2005, Exp. 1).
immediate and a delayed retention interval. The retrieval schedule did not modu-
late the difference between expanded and equal spaced retrieval and so we will
collapse across these schedules here. As shown in Figure 6.4, Logan obtained
benefits of expanded retrieval over equal interval during the acquisition phase for
both the young and the older adults. This acquisition advantage was retained on an
immediate final cued recall test, at least in older adults. However, after a 24 hour
delay between acquisition and final test, the advantage for items practiced on an
expanding schedule compared to an equal interval schedule was lost in both age
groups. In fact, memory was significantly worse for expanded items compared to
equal interval items in younger adults. Thus, the findings from Logan imply that
when an expanded retrieval advantage is found, it may be short-lived, compared to
comparable equal interval spacing.
The Logan results are particularly intriguing in light of a recent study by
Karpicke and Roediger (2005). These investigators compared massed (0-0-0),
equal interval (5-5-5), and expanded (1-5-9) retrieval conditions for paired associ-
ates (vocabulary GRE practice items, such as sobriquet–nickname) in both an
FIGURE 6.3 Mean final cued recall performance as a function of group and spacing
schedule from Balota et al. (2005, Exp. 1).
FIGURE 6.4 Mean final cued recall performance as a function of group, spacing sched-
ule, and retention interval from Logan (2004).
immediate 10 minute delayed, and 48 hour delayed cued recall test. The results
indicated that again both spaced schedules produced better performance than
the massed schedule at both retention intervals. More importantly, there was
evidence of a benefit of the expanded condition over the equal interval condition
in the immediate test, but the equal interval condition produced a clear benefit
in the delayed recall test. This later pattern was replicated in a second study in
which feedback was provided after each testing event, although the immediate
test performance was limited by ceiling problems. Hence, the short-term gains of
expanded over equal interval retrieval found in both the Logan (2004) and
Karpicke and Roediger (2005) studies were either eliminated or turned into
long-term losses in a delayed testing condition.
The results of the current review lead us to conclude that, as expected, spaced
practice produces considerable benefits in learning compared to massed practice;
however, the additional benefits of expanded practice over equal interval practice
have not been well substantiated in recent research. We find the lack of a benefit
of expanded retrieval quite intriguing, because the acquisition phase data (when
measured) clearly indicate that participants are at higher levels of retrieval success
over a longer retention interval in the expanded condition than in the equal inter-
val condition. For example, at the last retrieval event during the acquisition phase
of the Balota et al. (2005) study, subjects correctly retrieved 16% more items in
the expanded retrieval condition than in the equal interval condition, even though
at this last retrieval event, the expanded retention interval was five items, and the
equal interval was three items. Clearly, one would expect such a benefit to either
persist or increase during a final recall test. However, there was no benefit in the
final cued recall performance of expansion. Moreover, the results from Logan
(2004) and Karpicke and Roediger (2005) indicate that one can either eliminate or
actually reverse any acquisition benefit of expanded over equal interval retrieval
when tested at a 1 or 2 day retention interval.
Why might the benefits of expanded retrieval over equal interval retrieval
observed during the acquisition phase be lost in a later cued recall test? One
simple possibility is that long-term retention is simply a function of the average
amount of spacing. Since this is equated in the equal interval and expanded condi-
tion, there is, on average, no effect. A more intriguing possibility is that there are
counteracting influences, as a function of retention interval. Specifically, it is
indeed beneficial to maintain high levels of retrieval success during acquisition,
but there is a potential cost of maintaining these high levels, i.e., a loss of one of
the spaced intervals. Consider for example, the 1-3-5 schedule compared to the
3-3-3 schedule. The initial retrieval event during acquisition in the expanded
condition occurs after only one intervening item. In some sense, one could
argue that the 3-3-3 condition involves three functional spaced events during
acquisition, whereas, the 1-3-5 condition only involves two spaced events during
As noted earlier, one theoretical account of the spacing effect is encoding
variability theory, which nicely handles the intriguing spacing by retention interval
interaction. This framework may also be relevant to the expanded vs equal interval
spacing results. Specifically, although expanded retrieval does indeed produce
higher performance during the acquisition phase, this benefit may be lost in longer
delays, precisely as any benefit of massed spacing is relatively short lived. The
notion is that the equal interval condition will on average involve three distinct
encoding events, whereas, as noted above, the expanded condition on average
will involve two distinct encoding events. Because long-term retention is a reflec-
tion of the context at retrieval matching one of the encoding events, the equal
interval condition may produce a benefit in long-term retention, since this
condition affords an additional unique encoding of context. Interestingly, as
noted above, there is already evidence by Logan (2004) and Karpicke and
Roediger (2005) that initial benefits of expansion during a short retention inter-
val turn into losses a day or two later. Thus, although there is the benefit of
expansion in maintaining high levels of performance at increasing delays during
acquisition, the long-term consequence of expansion may produce a decreased
amount of contextual variability, because of the relatively immediate presenta-
tion of the first test. Of course, we know that maintaining high levels of retrieval
success during acquisition is not the only variable that is critical to memory
performance, since massed study produces the highest level of performance
during acquisition but the lowest level of performance during a long-term reten-
tion test. Clearly, further work is needed to better understand the balance
between spacing and retention interval in more complex expanded and equal
interval schedules.
Although the review of recent empirical work has questioned the benefits of
expanded retrieval over equal spaced retrieval, there are two reasons that may
support the use of expanded retrieval in more applied settings. First, as Camp and
colleagues have nicely demonstrated, this procedure is relatively easy to imple-
ment in a clinical setting. That is, gradually increasing retrieval intervals, while
maintaining success, benefits from feedback driven performance, i.e., providing
positive feedback at increasing retention intervals because of retrieval success.
Alternatively, a priori picking the best spaced retrieval practice schedule may
be relatively difficult to accomplish, especially for someone who is having global
cognitive impairments. Second, and more importantly, maintaining high levels of
retrieval success during acquisition in the face of increasing retention intervals is
likely to be reinforcing for the user. Of course, such reinforcement should ultim-
ately increase the likelihood that individuals will use such a schedule in the future,
and so this is a natural way of nurturing the use of spaced practice. Thus, although
there may be reasons to question the long-term benefits of expanded over equal
interval schedules, there may also be practical reasons to use the expanded retrieval
schedule to implement simple spaced acquisition.
Clearly, there is much work to do to better understand the influence of study
schedules on long-term retention. Comparing different acquisition schedules is a
natural avenue to pursue. Most studies have only used three retrieval attempts,
and the extension to a greater number of retrieval attempts would be likely to
reflect the sequence of events in more applied settings, such as in the Camp et al.
studies. Consider, for example, the possibility of comparing an expanded schedule
of 1-2-3-4-5-6-7 with a retrieval schedule of 4-4-4-4-4-4-4. Without feedback pro-
vided, we would expect that in this case the expanded schedule may produce a
benefit over the equal interval schedule, but this is an open empirical question.
Likewise, in addition to exploring different schedules, the influence of feedback
and the possibility of restudying nonretrieved information are both aspects of
acquisition that typically occur outside the laboratory. With feedback, one might
expect the equal interval condition to possibly produce better performance in the
above example. Finally, as Logan (2004) and Karpicke and Roediger (2005) have
nicely demonstrated, retention interval is critical. In general, what appears to be a
benefit of expanded retrieval during acquisition can be quickly lost and even
reversed at longer retention intervals. In this light, we are reminded of Bjork’s
(1999) arguments regarding the importance of desirable difficulties during acqui-
sition, and the counterintuitive observation that variables that produce benefits
immediately on tests sometimes produce losses in long-term retention. Although
there will clearly be constraints on when expansion will and will not produce
benefits over equal interval schedules, the present review makes clear that the
power of spacing is paramount in learning and memory.
Portions of this research were supported by NIA PO1 AGO3991 and P50 AGO5681.
Thanks are extended to Roddy and Jim Nairne for helpful comments on an earlier version
of this manuscript.
Baddeley, A. D., & Wilson, B. A. (1994). When implicit learning fails: Amnesia and the
problem of error elimination. Neuropsychologia, 32, 53–68.
Balota, D. A., Duchek, J. M., & Paullin, R. (1989). Age-related differences in the impact of
spacing, lag and retention interval. Psychology and Aging, 4, 3–9.
Balota, D. A., Duchek, J. M., Sergent-Marshall, S. D., Roediger, H. L., III. (2005). Does
expanded retrieval produce benefits over equal interval spacing? Explorations of
spacing effects in healthy aging and early stage Alzheimer’s disease. Psychology and
Aging, 21, 19–31.
Balota, D. A., & Ferraro, F. R. (1996). Lexical, sublexical, and implicit memory processes
in healthy young, healthy older adults, and in individuals with dementia of the
Alzheimer’s type. Neuropsychology, 10, 82–95.
Bjork, R. A. (1999). Assessing our own competence: Heuristics and illusions. In D. Gopher
& A. Koriat (Eds.), Attention and performance XVII: Cognitive regulation of per-
formance: Interaction of theory and application (pp. 435–459). Cambridge, MA:
MIT Press.
Bourgeois, M. S., Camp, C., Rose, M., White, B., Malone, M., Carr, J., & Rovine, M.
(2003). A comparison of training strategies to enhance use of external aids by
persons with dementia. Journal of Communication Disorders, 36, 361–378.
Brush, J. A., & Camp, C. J. (1998a). Using spaced retrieval to treat dysphagia on a long-term
care resident with dementia. The Clinical Gerontologist, 19(2), 96–99.
Brush, J. A., & Camp, C. J. (1998b). Using spaced retrieval as an intervention during
speech-language therapy. The Clinical Gerontologist, 19, 51–64.
Burke, D. M., & Light, L. L. (1981). Memory and aging: The role of retrieval processes.
Psychological Bulletin, 90, 513–546.
Camp, C. J. (1989). Facilitation of new learning in Alzheimer’s disease. In G. C. Gilmore,
P. J. Whitehouse, & M. L. Wykle (Eds.), Memory, aging, and dementia
(pp. 212–225). New York: Springer.
Camp, C. J., Bird, M. J., & Cherry, K. E. (2000). Retrieval strategies as a rehabilitation
aid for cognitive loss in pathological aging. In R. D. Hill, L. Backman, A. Neely
Stigsdotter (Eds.), Cognitive rehabilitation in old age (pp. 224–248). Oxford, UK:
Oxford University Press.
Camp, C. J., Foss, J. W., Stevens, A. B., & O’Hanlon, A. M. (1996). Improving prospective
memory task performance in persons with Alzheimer’s disease. In M. Bandimonte,
G. O. Einstein, & M. A. McDaniel (Eds.), Prospective memory: Theory and
applications (pp. 351–367). Mahwah, NJ: USum Associates.
Camp, C. J., & Mattern, J. M. (1999). Innovations in managing Alzheimer’s disease. In
D. E. Biegel & A. Blum (Eds.), Innovations in practice and service delivery across
the lifespan (pp. 276–293). New York: Oxford University Press.
Camp, C. J., & McKitrick, L. A. (1992). Memory interventions in Alzheimer’s-type
dementia populations: Methodological and theoretical issues. In R. L. West & J. D.
Sinnott (Eds.), Everyday memory and aging: Current research and methodology
(pp. 152–172). New York: Springer-Verlag.
Camp, C. J., & Schaller, J. R. (1989). Epilogue: Spaced-retrieval memory training in an
adult day care center. Educational Gerontology, 15, 81–88.
Camp, C. J., & Stevens, A. B. (1990). Spaced-retrieval: A memory intervention
for dementia of the Alzheimer’s type. The Clinical Gerontologist, 10(1),
Carpenter, S. K., & DeLosh, E. L. (2005). Application of the testing and spacing effects to
name learning. Applied Cognitive Psychology, 19, 619–636.
Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice
in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin,
132, 354–380.
Cherry, K. E., Simmons, S. S., & Camp, C. J. (1999). Spaced retrieval enhances memory in
older adults with probable Alzheimer’s disease. Journal of Clinical Geropsychology,
5(3), 159–175.
Clare, L., Wilson, B. A., Carter, G., Breen, K., Gosses, A., & Hodges, J. R. (2000). Interven-
ing with everyday memory problems in dementia of Alzheimer type: An errorless
learning approach. Journal of Clinical and Experimental Neuropsychology, 22,
Crowder, R. G. (1976). The principles of learning and memory. Oxford, UK. Lawrence
Erlbaum Associates Ltd.
Cull, W. L. (2000). Untangling the benefits of multiple study opportunities and repeated
testing for cued recall. Applied Cognitive Psychology, 14, 215–235.
Cull, W. L., Shaughnessy, J. J., & Zechmeister, E. B. (1996). Expanding understanding of
the expanding-pattern-of-retrieval mnemonic: Toward confidence in applicability.
Journal of Experimental Psychology: Applied, 2(24), 365–378.
Dempster, F. N. (1996). Distributing and managing the conditions of encoding and prac-
tice. In E. L. Bjork & R. A. Bjork (Eds.), Handbook of perception and cognition:
Memory (pp. 317–344). San Diego, CA: Academic Press.
Duchek, J. M. (1984). Encoding and retrieval differences between young and old: The
impact of attentional capacity usage. Developmental Psychology, 20(6), 1173–1180.
Estes, W. K. (1955). Statistical theory of distributional phenomena in learning.
Psychological Review, 62, 369–377.
Faust, M. E., Balota, D. A., & Spieler, D. H. (2001). Building episodic connections: Changes
in episodic priming with age and dementia. Neuropsychology, 15, 626–637.
Glenberg, A. M. (1977). Influences of retrieval processes on the spacing effect in free recall.
Journal of Experimental Psychology: Human Learning and Memory, 3(3), 282–294.
Greene, R. L. (1992). Repetition paradigms. In R. L. Greene (Ed.), Human memory:
Paradigms and paradoxes (pp. 132–152). Hillsdale, NJ: Lawrence Erlbaum
Associates, Inc.
Hayden, C. M., & Camp, C. J. (1995). Spaced-retrieval: A memory intervention for
dementia in Parkinson’s disease. The Clinical Gerontologist, 16(3), 80–82.
Hochhalter, A. K., Overmier, J. B., Gasper, S. M., Bakke, B. L., & Holub, R. J. (2005). A
comparison of spaced retrieval to other schedules of practice for people with
dementia. Experimental Aging Research, 31, 101–118.
Karpicke, J. D. (2004). Test-enhanced learning: The effects of repeated tests and the
spacing of tests on long-term retention. Unpublished Masters thesis, Washington
University, St. Louis, MO.
Karpicke, J. D., & Roediger, H. L., III. (2005). Does expanding retrieval work? Poster
presented at the Midwestern Psychological Association conference, Chicago.
Landauer, T. K., & Bjork, R. A. (1978). Optimum rehearsal patterns and name learning.
In M. Gruneberg, P. E. Morris, & R. N. Sykes (Eds.), Practical aspects of memory
(pp. 625–632). London: Academic Press.
Lee, M. M., & Camp, C. J. (2001). Spaced-retrieval: A memory intervention for HIV+ older
adults. The Clinical Gerontologist, 22(3/4), 131–135.
Logan, J. M. (2004). Spaced and expanded retrieval effects in younger and older adults.
Unpublished doctoral dissertation, Washington University, St. Louis, MO.
McGeogh, J. A. (1943). The psychology of human learning. New York: Longmans Green.
McKitrick, L. A., & Camp, C. J. (1993). Relearning the names of things: The spaced-
retrieval intervention implemented by a caregiver. The Clinical Gerontologist,
14, 60–62.
McKitrick, L. A., Camp, C. J., & Black, F. W. (1992). Prospective memory intervention in
Alzheimer’s disease. Journal of Gerontology: Psychological Sciences, 47, 337–343.
Melton, A. W. (1970). The situation with respect to the spacing of repetitions and memory.
Journal of Verbal Learning and Verbal Behavior, 9, 596–606.
Morris, P. E., & Fritz, C. O. (2000). The name game: Using retrieval practice to improve the
learning of names. Journal of Experimental Psychology: Applied, 6, 124–129.
Morris, P. E., & Fritz, C. O. (2002). The improved name game: Better use of expanding
retrieval practice. Memory, 10, 259–266.
Morris, P. E., Fritz, C. O., & Buck, S. (2004). The name game: Acceptability, bonus
information and group size. Applied Cognitive Psychology, 18, 89–104.
Peterson, L. R., Wampler, R., Kirkpatrick, M., & Saltzman, D. (1963). Effect of spacing
presentations on retention of a paired associate over short intervals. Journal of
Experimental Psychology, 66(2), 206–209.
Rea, C. P., & Modigliani, V. (1985). The effect of expanded versus massed practice on the
retention of multiplication facts and spelling lists. Human Learning: Journal of
Practical Research and Applications, 4(1), 11–18.
Riley, K. P. (1992). Bridging the gap between researchers and clinicians: Methodological
perspectives and choices. In R. L. West & J. D. Sinnott (Eds.), Everyday
memory and aging: Current research and methodology (pp. 182–189). New York:
Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory
tests improves long term memory. Psychological Science, 17, 249–255.
Schacter, D. L., Rich, S. A., & Stampp, M. S. (1985). Remediation of memory disorders:
Experimental evaluation of the spaced-retrieval technique. Journal of Clinical and
Experimental Neuropsychology, 7, 70–96.
Spieler, D. H., & Balota, D. A. (1996). Characteristics of associative learning in younger and
older adults: Evidence from an episodic priming paradigm. Psychology and Aging,
11, 607–620.
Spitzer, H. F. (1939). Studies in retention. Journal of Educational Psychology, 30, 641–657.
Wilson, B. A., Baddeley, A., Evans, J., & Shiel, A. (1994). Errorless learning in the rehabili-
tation of memory impaired people. Neuropsychological Rehabilitation, 4, 307–326.
Wilson, B. A., & Evans, J. J. (1996). Error-free learning in the rehabilitation of people with
memory impairments. Journal of Head Trauma Rehabilitation, 11, 54–64.
... Most of the studies on retrieval schedules compared expanding and massed retrieval, but did not include the critical equal interval condition needed to compare expanding retrieval to another distributed schedule (e.g., Rea & Modigliani, 1985). All studies making the massed versus expanding retrieval comparison showed expanding retrieval to be more effective, and Balota, Duchek, and Logan (2007) have provided an excellent review of this literature. They show conclusively that massed testing is a poor strategy relative to distributed testing, despite the fact that massed testing produces very high performance on the initial tests (much higher than equal interval testing). ...
... On the other hand, when comparisons are made between expanding and equal interval schedules, the data are much less conclusive. The other main point established in the Balota et al. (2007) review is that no consistent evidence exists for the advantage of expanding retrieval schedules over equal interval testing sequences. A few studies after Landauer and Bjork's (1978) seminal study obtained the effect, but the great majority did not. ...
... However, after a 2-day delay, recall was best in the equal interval condition relative to the expanding condition, although both conditions still produced better performance than in the massed condition The conclusion that retrieval difficulty is the key element promoting better long-term retention for equal interval (relative to expanding or massed) schedules fits well with data reviewed in previous parts of the chapter, such as recall tests producing a greater testing effect than recognition tests. However, as Balota et al. (2007) have noted, the literature on expanding retrieval sequences has used a relatively small array of conditions (three tests, paired associate learning of one sort or another, relatively immediate tests). One can imagine that an expanding retrieval schedule could be better than equal intervals if there were more presentations and these occurred over longer periods of time. ...
... The study-phase retrieval theory posits that when an item is repeated for study, the recognition of that item triggers the retrieval of the prior study experience for that item Hintzman et al., 1975). This manifestation of retrieval-practice benefits (Bjork, 1988;Roediger III & Butler, 2011; see also Balota et al., 2007a) becomes even greater/stronger when the eventual later successful retrieval of an item is more difficult or delayed (Appleton-Knapp et al., 2005;Greene, 1989;Thios & D'Agostino, 1976 Benjamin & Tullis, 2010), which may be why the benefits of spacing are so prevalent and general. ...
... Specifically, the cumulative rehearsal occurring when only given a single (but longer) opportunity to study each item may be replaced with more effective encoding strategies on later presentations. There may also be retrieval practice benefits (Bjork, 1988;Roediger III & Butler, 2011; see also Balota et al., 2007a) when words are presented multiple times, leading to enhanced recall, particularly for words in the middle of the list (as evidenced by the poorer recall of mid-list words when only studying an item a single time). ...
Full-text available
We assessed the effects of removing some constraints that characterize traditional experiments on the effects of spaced, rather than massed, study opportunities. In five experiments—using lists of to-be-remembered words—we examined the effects of how total study time was distributed across multiple repetitions of a given to-be-remembered word. Overall, within a given list, recall profited from study time being distributed (e.g., four 1-second presentations or two 2-second presentations versus one 4-second presentation). Among the implications of these findings is that if students choose to engage in massed studying (by virtue of constraints on their study time or a failure to appreciate the benefits of spaced study sessions), then studying the information twice but for half the time may produce memory benefits in a single study session.
... This approach of repetitive assessment has been shown to be an effective mechanism to improve learning, specifically in introductory biology courses (38,39). Another area of significant research has shown that the "spacing effect, " whereby the repetition of the material to be learned is spaced, leads to increased learning and memory for a large variety of content areas (40,41). The intervals shown to be effective for spaced learning, including for life science curricula (42), are on the order of the spacing between repetitive assessments of calculations employed here. ...
... Indeed, this has been shown to be an effective approach for introductory biology courses (38,39). The intervals between assessments used in this study are also consistent with mechanisms of the "spacing effect" (40)(41)(42). For example, a study that assessed the spacing effect for natural science material in a lecture context showed that spaced review 8 days after presentation of material led to increased learning gains as opposed to review 1 day after (42). ...
Full-text available
ABSTRACT Quantitative reasoning is one of the core competencies identified as a priority for transforming the undergraduate biology curriculum. However, first-year biology majors often lack confidence in their quantitative skills. We revised an introductory biology lab to emphasize the teaching of basic laboratory calculations, utilizing multiple teaching tools, including online prelab quizzes, minilab lectures, calculation worksheets, and online video tutorials. In addition, we implemented a repetitive assessment approach whereby three types of basic calculations—unit conversions, calculating molar concentrations, and calculating dilutions—were assessed on all quizzes and exams throughout the semester. The results showed that learning improved for each of the three quantitative problem types assessed and that these learning gains were statistically significant, both from first assessment to midterm and, notably, from midterm to final. Additionally, the most challenging problem type for students, calculating molar concentrations, showed the greatest normalized learning gains in the second half of the semester. The latter result suggests that persistent assessment resulted in continued learning even after formal, in-class teaching of these approaches had ended. This approach can easily be applied to other courses in the curriculum and, given the learning gains achieved, could provide a powerful means to target other quantitative skills.
... Previous research suggests that retrieval practice, the act of spacing tests throughout the learning process, can be helpful for improving memory (for reviews, see Balota, Duchek, & Logan, 2007;Cepeda, Pashler, Vul, Wixted, & Rohrer, 2006). However, the amount of time between tests is also important, as some research suggests that younger adults benefit more from a short lag between tests relative to a longer lag (Maddox & Balota, 2015). ...
Contending with talker variability has been found to lead to processing costs but also benefits by focusing learners on invariant properties of the signal, indicating that talker variability acts as a desirable difficulty. That is, talker variability may lead to initial costs followed by long‐term benefits for retention and generalization. Adult participants learned an artificial grammar affording learning of multiple components in two experiments varying in difficulty. They learned from one, two, or eight talkers and were tested at three time points. The eight‐talker condition did not impact learning. The two‐talker condition negatively impacted some aspects of learning, but only under more difficult conditions. Generalization of the grammatical dependency was difficult. Thus, we discovered that high and limited talker variability can differentially impact artificial grammar learning. However, talker variability did not act as a desirable difficulty in the current paradigm as the few evidenced costs were not related to long‐term benefits.
... For example, in arranging tests as recall of word lists, paired words (i.e., foot-shoe, or nonsense-words as, GRATspoon), or prose (to simulate the material commonly used in college settings), many laboratory studies showed consistently better performance on retention for the participants who had the opportunity to be exposed to tests (in form of self-recitation, elaboration of speculative questions or scenarios, multiple-choice questions, or short open answers) as compared to the participants who were required to only study the text. Multiple but spaced-in-time testing opportunities determined better results too, in comparison with a single test exposure, or massive opportunities crammed into a few sessions (Balota et al., 2007;Bangert-Drowns et al., 1991;Karpicke & Roediger, 2010;Tulvin, 1967;Wheeler & Roediger, 1992). Glodowski and colleagues (2019) present several evidence-based teaching recommendations elaborated from an extensive review of field studies involving the use of tests with college students. ...
Full-text available
Modern societies need higher education systems that are strongly grounded on scientific knowledge and evidence-based teaching tools. This laboratory-based study extended Chase and Houmanfar (2009), examining the effects of basic and elaborate feedback on participants’ performance measured with two sets of tests administered after a college-level text was studied. The use of Tobii® Pro X2-60 eye trackers allowed for detection of students’ orientation toward the displayed feedback. The eye tracker results showed that participants spent more time reading their incorrect answers than the correct answers, but this lingering of their eye gazes on the screen did not have a positive correlation with the performance and the final test. However, other factors such as the feedback provision need to be more closely considered. The experimental findings also demonstrated that the presence of feedback following the first test phase improved the performance of the final test and that the elaborate feedback had the largest effect when the participants where naïve about the topic studied during the experimental session.
... As an important exception, some recent procedures use the same set of stimuli to allow direct comparison between working and long-term memory. Factors that may be considered confounds by working memory theorists are of great interest in the context of longterm learning, such as repeated exposure (Slamecka & McElree, 1983) or repeated testing of information (Balota, Duchek, & Logan, 2007). However, the interaction of working-and long-term memory systems has crucial implications for learning. ...
In this chapter, we discuss the crucial role of working memory limitations in long-term learning. Convincing correlational evidence suggests that working memory ability is related to various measures of educational attainment. Here, we focus on how and why reduced working memory ability may restrict long-term learning. Specifically, we consider the working memory system as a bottleneck for long-term learning, constraining the ability to learn the meaning of new concepts and encode new information. We also discuss how working memory plays a direct role in a learner's ability to follow instructions and stay on task. Then, we discuss the essential role that long-term memory knowledge plays in ongoing working memory processing. For instance, prior knowledge allows us to categorize and chunk information. The interaction between working memory and long-term memory may form a positive feedback loop with development. For example, as children develop, better working memory might allow more knowledge accumulation, which could, in turn, help in the offloading of working memory to long-term memory, leaving more working memory free to facilitate subsequent learning activities. We suggest that adjusting learning materials according to the learners' working memory abilities—and long-term memory knowledge—may allow the most efficient learning.
... The second open debate is on the fixed versus variable intervals (expanding intervals). The research on this problem has yielded ambiguous results (Balota et al., 2007;Logan & Balota, 2008). ...
Full-text available
Training design for automatic skills has a vast domain of application, such as education, physical and cognitive rehabilitation, as well as sports, arts and professional training. Gamification concept used in technology-assisted training has the potential to increase motivation, engagement and adherence to the training programme. Currently, the general gamification models of learning, did not take into account the temporal specificity of the game elements for automaticity acquisition training. In order to address this problem, an extensive overview of the key training attributes that impact automaticity acquisition was carried out. Then, based on this review, the three steps of a proposed model were presented. The first step of this model, named Task Analytics, helps with task-specific training decisions. The second step provides descriptive and prescriptive approaches for the three phases of automaticity acquisition (fast learning, slow learning and automatization). The descriptive part characterizes each phase using psychological and performance-related qualities, while the prescriptive part recommends the appropriate training elements for each phase. Based on the prescriptive part, a game-design model is proposed in the third step, which classifies the game mechanics and maps them onto each phase of automaticity acquisition. Finally, to validate this approach, a mobile game was designed based on the proposed gamification model, and it was compared to control design. The two approaches are tested with 49 participants. The results showed that the experimental group had a significantly better engagement and higher performance. Furthermore, the experimental group showed significantly better performance in a multitasking challenge designed to evaluate the automaticity. The main contribution of this article is the proposed game design model that takes into account the temporal specificity of game elements during the acquisition of automaticity.
Dyslexia is the most prevalent learning disability and the lack of automaticity in word recognition is one of its main characteristics. This thesis is dedicated to the gamification of in-home dyslexia remediation by focusing on the development of automaticity for children aged from five to nine years old. An extensive research on the effective principles of automaticity training, led to the creation of a novel training model for automaticity acquisition. Training for automaticity requires countless repetitions, and adherence is crucial. Gamification concept was proposed to increase motivation, engagement, and adherence. Based on the automaticity-training model, a gamification model was developed for placing each game element at its appropriate phase of automaticity acquisition. The effectiveness of this gamification model was validated by a randomized controlled trial. Furthermore, an optimization model was developed to provide individualized training sessions based on the level of the learner. For estimating the difficulty level of the content, four lexical skills were modelled using artificial neural networks and linear regression. For each lexical skill, the top 10 lexical variables were identified through forward stepwise analysis. The accuracy of the models based on mean absolute error reached 90.58% for auditory word recognition, 92.08% for visual word recognition, 84.83% for spelling, and 86.98% for word decoding. Finally, based on these developments, four games were created and provided on multiple platforms. The evaluation of the usability study confirmed the viability of the intelligent system and the games scored 80% on the system usability scale.
Full-text available
The main goal of this research is to systemize, build, and test prototype software to demonstrate Indirect Spaced Repetition (ISR) as a viable concept for Second Language Vocabulary Acquisition (SLVA). ISR is designed around well-founded spaced repetition and SLVA principles. Most importantly, it is based on Nation’s (2001) recommendation to consider all three tiers of word knowledge (meaning, form, and function/use) and subsequent 18 aspects of word knowledge for a more balanced approach in teaching and learning vocabulary. ISR prototype software was achieved in the conceptual phase of the research. The resulting prototype flashcard software was given an in-depth trial for a period of 2 weeks by seven university students. Participants were given a post-project survey to evaluate ISR software (ISRS) under four categories: enjoyment, usefulness, usability, and general consideration. Post-test survey findings showed above-average satisfaction and consideration to use such software in the future. However, these findings also revealed that some areas could be further improved, such as addressing some hardware/software issues (e.g., IT infrastructure problematics and lag) and integrating gamification elements (e.g., performance feedback/reports).
Although the benefits of spaced retrieval are well established, research suggests that young and older adults often fail to optimally implement this strategy. The present study examined how task experience with feedback influenced participant-implemented spaced retrieval and its effect on short and long-term memory retention. Young and older adults were instructed to either equally space or expand their retrieval of face-name associations throughout an ongoing reading task. Participants were then provided feedback on the accuracy with which they implemented experimenter instructions. Results showed that feedback improved utilization of retrieval practice in both young and older adults. Moreover, both age groups successfully produced a pattern of expanded retrieval when instructed to do so, but failed to properly implement equal spacing. Consistent with extant research utilizing experimenter-determined spaced retrieval schedules, our study showed that the inclusion of a longer spacing interval immediately following acquisition resulted in reduced forgetting across the retention interval.
Dementia of Alzheimer Type (DAT) is increasingly detected at an earlier stage of the disorder, when interventions to assist with everyday memory difficulties might be most valuable. Some learning is possible in DAT and a number of factors have been identified which may facilitate performance, although applications to everyday memory problems have been limited. The concept of errorless learning has not previously been directly examined in relation to DAT, but might provide a useful additional strategy. In the present study, 6 participants with early stage DAT (MMSE scores 21 - 26) received individually tailored interventions, based on errorless learning principles and targeted at a specific everyday memory problem. Five of the participants showed significant improvement on the target measures, and maintained this improvement up to 6 months later. The results suggest that it is feasible to intervene with everyday memory problems in the early stages
The emergence of everyday memory and aging as a major field of study has resulted from the combined efforts of investigators from a variety of backgrounds, interests, and disciplines, including experimental researchers as well as practicing clinicians. While the findings of everyday memory research are clearly relevant to the development of intervention programs, efforts to implement laboratory-based research methods and designs in clinical settings are still in the relatively early stages of development. This chapter will discuss the contributions that can be made by practicing clinicians to the field of everyday memory and aging, focusing on two themes. First, the methodological choices made by experimental researchers will be considered in terms of their utility and relevance to the practitioner who wishes to apply these methods in training programs with clinical populations. A second issue to be discussed is the role of the clinician in initiating interventions that may then be examined on a larger scale by experimental researchers in the field of everyday memory and aging. A small number of recent studies on memory skills training programs will be used to illustrate these themes, emphasizing my work and the work of others with persons who have been diagnosed as memory impaired. The interested reader is referred to recent reviews of the literature on memory-training programs for a more detailed discussion of findings in this area (West, 1990; West & Tomer, 1989; see also Chapter 10, this volume).
In this chapter, we will consider methodological and theoretical issues encountered while implementing memory interventions in individuals with Alzheimer’s-type Dementia (AD). The severe memory impairments associated with AD have profound impact on the ability of afflicted individuals to carry out daily activities, and any new learning (or relearning) tends to be limited and specific. Interventions in this population must address the practical needs of individuals as they cope with a reality configured by impairments in everyday memory (i. e., memory tasks encountered in the real world). Previous attempts to modify the memorial abilities of individuals with AD have met with little success, when they have been tried at all, and we will discuss the problems associated with using traditional mnemonic training with this population. We will then describe a relatively new intervention—spaced retrieval—which seems to overcome many of these difficulties. Results from recent studies using this technique with individuals having AD will be reported.
This chapter reviews interventions for managing Alzheimer's disease (AD). It begins with an overview of AD, including its stages and risk factors. It then discusses innovative interventions including spaced-retrieval focused on individuals with AD; question-asking reading, an intervention for small groups of persons with AD; and an Intergenerational Montessori-based program for dyads of persons with AD and preschool children.
Spacing multiple study opportunities apart from one another is known by psychologists to be a highly effective study method (see Dempster, 1996). This study examines whether including tests during study would produce practical benefits for learning beyond that provided by distributed study alone. In addition, spacing of both study and test (massed, uniform distributed, and expanding distributed) is investigated. To-be-remembered information was repeated with a single learning session (Experiment 1), reviewed immediately after initial learning (Experiment 2), or reviewed days after initial learning (Experiments 3 and 4). As expected, large distributed practice effects were shown across experiments. In addition to these effects, testing produced significant benefits for learning in all four experiments, which were of moderate or large size (Cohen's d of 0.52 to 1.30) for three experiments. Expanding test spacing, however, did not independently benefit learning in any of the learning situations studied. Educators should take advantage of the large benefits that distributed study and testing have on learning by spacing multiple tests of information within learning sessions and by distributing tests across multiple review sessions.
A memory‐impaired client at an adult day‐care center was trained to remember the name of a staff member using the spaced‐retrieval technique. After two brief training sessions, the client demonstrated long‐term retention of the target name. Retention was maintained at follow‐up testing 6 months later. Details of the unique aspects of this cognitive intervention, along with implications of this project for cognitive aging research in general, are discussed. In particular, the social and emotional components of memory‐ and cognitive training tasks are emphasized.