Archives of Physical Medicine and Rehabilitation
Volume 89, Issue 8 , Pages 1514-1521, August 2008

Self-Generation to Improve Learning and Memory of Functional Activities in Persons With Multiple Sclerosis: Meal Preparation and Managing Finances

Presented to the Consortium of Multiple Sclerosis Center, June 1–5, 2005, Orlando, FL, and the International Neuropsychological Society, February 2–5, 2005, St. Louis, MO.

  • Yael Goverover, PhD, OT

      Affiliations

    • Kessler Medical Rehabilitation Research and Education Center, West Orange, NJ
    • Department of Occupational Therapy, Steinhardt School of Culture, Education, and Human Development, New York University, New York, NY.
    • ,
  • Nancy Chiaravalloti, PhD

      Affiliations

    • Kessler Medical Rehabilitation Research and Education Center, West Orange, NJ
    • University of Medicine and Dentistry of New Jersey/New Jersey Medical School, Newark, NJ
    • ,
  • John DeLuca, PhD, ABPP

      Affiliations

    • Kessler Medical Rehabilitation Research and Education Center, West Orange, NJ
    • University of Medicine and Dentistry of New Jersey/New Jersey Medical School, Newark, NJ
    • Corresponding Author InformationReprint requests to John DeLuca, PhD, ABPP, Kessler Medical Rehabilitation Research and Education Center, Neuropsychology and Neuroscience Laboratory, 300 Executive Dr, Ste 010, West Orange, NJ 07052

published online 01 July 2008.

Article Outline

Abstract 

Goverover Y, Chiaravalloti N, DeLuca J. Self-generation to improve learning and memory of functional activities in persons with multiple sclerosis: meal preparation and managing finances.

Objective

To examine the utility of using a self-generation strategy to improve learning and performance of everyday functional tasks in persons with multiple sclerosis (MS).

Design

Mixed-design with both a within- and between-subject factor.

Setting

Nonprofit rehabilitation research institution.

Participants

Participants (n=20) with MS and healthy controls (n=18).

Interventions

Participants completed 2 meal preparation and 2 financial management tasks. One task in each area was presented in the provided condition, in which all instructions were provided to and read by the participants, and the other task was presented in the generated condition, in which participants were asked to generate (fill in the blank) the necessary items needed to perform each step of the task.

Main Outcome Measures

Correct recall of task items and step sequence immediately and 1 week after initial learning and correct performance of task items and step sequence 30 minutes after initial learning. The maximum possible score in each of the recall tests was 24.

Results

Although the MS and healthy groups did not differ in overall items recalled, in both groups tasks learned in the generated condition enhanced memory performance significantly for the tasks used when compared with similar tasks learned in the provided condition.

Conclusions

Self-generation during learning can significantly improve subsequent recall of information and performance of activities of daily living for persons with MS. Implications of these findings for cognitive rehabilitation in MS are discussed.

Key Words: Activities of daily living, Learning, Memory, Multiple sclerosis, Rehabilitation

List of Abbreviations: AMPS, Assessment of Motor and Process Skills, ANOVA, analysis of variance, BNT, Boston Naming Test, CMDI, Chicago Multiscale Depression Inventory, CVLT, California Verbal Learning Test, D-KEFS, Delis-Kaplan Executive Function System, IADLs, instrumental activities of daily living, MS, multiple sclerosis, SDMT, Symbol Digit Modalities Test, STAI, State-Trait Anxiety Inventory, TMT, Trail-Making Test, WAIS-R, Weschler Adult Intelligence Scale−Revised

 

IT IS NOW WELL ESTABLISHED that up to two thirds of persons with MS have important cognitive difficulties.1, 2 These difficulties have been shown to directly influence everyday-life functional activities, including one's ability to maintain employment.3, 4, 5 The most prominent cognitive deficits in persons with MS involve learning and memory.6, 7, 8 Studies have shown that the primary reason for difficulties in learning and memory in persons with MS is difficulty in the initial acquisition of information, rather than the retrieval of information from long-term storage in the brain.6, 7, 9, 10 DeLuca,6, 7 Gaudino,8 Demaree,9 and colleagues showed that when persons with MS were equated for the amount of information initially acquired on a list-learning task by reaching a common learning criterion, recall and recognition performance did not significantly differ from that of healthy controls at 30-minute, 90-minute, and 1-week delays after learning. However, participants with MS needed significantly more trials to reach the learning criterion. That is, they had difficulty initially learning new information, but not in retrieval from long-term storage. This series of research studies shows that learning is a major source of the episodic memory deficit often observed in persons with MS. Given that learning is the key problem in persons with MS, then treatment focusing on improving learning may increase recall and recognition, which in turn can help improve performance of daily activities.

Self-generation is a learning strategy where persons are asked to generate their own words, concepts, or items to improve their learning and memory. Decades of research in cognitive psychology have shown that items self-generated by a person, such as words or concepts, are better remembered than items simply read or heard (ie, provided).11, 12 This phenomenon, known as the generation effect, has proven to be robust in improving initial acquisition and learning as well as subsequent recall and recognition of new materials in healthy persons.12, 13, 14 Theoretically, self-generation enhances encoding through deep processing of information (ie, more elaborative and personal relevance), and therefore this information can be retrieved more efficiently.13, 15

Although the generation effect has been widely studied in healthy participants, we have found few studies that have applied the self-generation strategy to participants with neurologic illness or injury. Such participant populations have included traumatic brain injury,16, 17 epilepsy,18 Parkinson's disease,19 and dementia,20, 21, 22, 23 with findings indicating that items learned in generated condition are remembered better than items learned in a provided condition. Chiaravalloti and DeLuca24 were the first to investigate the effect of self-generation of words in participants diagnosed with MS. They showed that using a self-generation strategy significantly increased recall and recognition on a laboratory-based task in persons with MS. However, after a 1-week delay, the degree of benefit from self-generation decreased significantly and participants recalled the same amount of information regardless of the learning condition.

The above studies documenting the benefit obtained by self-generation typically used laboratory-based stimuli and tasks such as words,13, 23 numbers,25 multiplication tasks,26 or pictures.27, 28 Although such studies are important to establish the effectiveness of self-generation for improving learning and memory in neurologic populations, it is unclear from laboratory-based tasks alone whether similar benefits can be actually obtained when attempting to improve a person's everyday memory and everyday life. To our knowledge, 1 study29 has investigated the application of self-generation to learning information related to daily living in persons with MS (eg, people's names, object locations, appointments). Results indicated that encoding through self-generation increased memory for this information, but a much stronger effect was observed for the appointment and object location tasks than for the name learning tasks.

The purpose of the present study is to examine the utility of self-generation strategy in improving recall and performance of actual everyday tasks, specifically; meal preparation (eg, preparing breakfast foods) and managing finances (eg, pay an electric bill) in persons with MS. It was hypothesized that participants will show better recall of functional everyday life tasks presented in the generated condition as compared with tasks presented in the provided condition. Similarly, it was hypothesized that the actual performance of the task will be better when tasks are learned in the generated condition as compared with the provided condition. Last, we examined whether type of task (ie, meal preparation vs managing finances) will have an effect on the benefits of self-generation.

Back to Article Outline

Methods 

Participants 

Participants consisted of 20 persons with clinically definite MS, diagnosed according to the criteria of Poser et al,30 and 18 healthy control participants without reported neurologic disabilities. We excluded participants from the present study if they had: (1) a history of neurologic illness or injury (aside from MS), (2) a history of major psychiatric illness, (3) a history of alcohol or drug abuse, or (4) severe visual or motor impairment that might interfere with the study procedures (ie, testing and performance of study tasks). All MS participants were at least 1 month after most recent exacerbation and were free of corticosteroid use. The 2 groups did not differ significantly with regard to age, education, sex, and premorbid intelligence (measured by Wide Range Achievement Test–3, reading subtest31) (table 1). With regard to disease type, 80% (n=16) of the participants with MS had a relapsing remitting course, 10% (n=2) of the participants had a primary progressive course of MS, and 10% (n=2) had secondary progressive MS.

Table 1. Demographic Data
CharacteristicMS Group (n=20)Control Group (n=18)FP
Age (y)46.1±8.840.4±11.92.7.10
Education (y)15.5±2.516.8±1.53.9.06
WRAT-3 score49.2±4.551.6±3.43.3.08
Sex, % (n) =.32.72
Males25(5)33.3(6)
Females75(15)66.7(12)
Disease type, % (n)
Relapsing remitting80(16)NA
Primary progressive10(2)NA
Secondary progressive10(2)NA
Disease duration (y)12.1±6.8NA

NOTE. Values are mean ± SD or as otherwise indicated.

Abbreviations: NA, not applicable; WRAT-3, Wide Range Achievement Test–3.

Materials 

Generation effect protocol: tasks 

We presented participants with 4 IADL tasks in each of 2 categories: 1 category related to meal preparation and the other related to managing finance. Each category contained 2 tasks (ie, 2 cooking tasks, 2 managing finance tasks). The tasks in this study were (1) meal preparation (including preparing French toast and preparing an omelet and toast) and (2) managing finance (including paying an electrical bill and ordering clothing from a catalog). The meal preparation tasks were taken from the AMPS.32 Tasks on the AMPS are calibrated according to level of difficulty. For the present study, the 2 meal preparation tasks had the same level of difficulty based on the AMPS. The managing finance tasks were developed specifically for the present study, and were equated for level of difficulty by task analysis conducted by 2 occupational therapists. Each of the tasks related to managing finances was analyzed separately by an occupational therapist. Each occupational therapist divided the task into steps and described what motor and process skills are needed for successful completion of each of the tasks. Finally, the 2 separate analyses, of each task, were compared for equivalency.

Each of the 4 tasks was subdivided into 12 individual steps, with each step presented individually on 10×23-cm index cards. Scores were measured on 2 dimensions: (1) recall or performance of each of the 12 steps required to complete the task and (2) the sequence of 12 task step recall or performance. Thus, 2 points were scored for every step, 1 point for remembering the item and 1 point for the correct sequence of the step. Scores for each task varied from 0 (was not able to remember any step) to 24 (correct recall and correct sequence).

Generation effect protocol: conditions 

Within each category (ie, meal preparation, managing finance), 1 task was presented in the generated condition and the other in the provided condition. In the provided condition a set of 12 completed steps were presented. The instructions were as follows: “You are going to read aloud steps (in a recipe or how to perform a managing finance task) one at a time. Try to remember the steps because I will ask you about them later.”

In the generated condition the 12 steps were presented with the last or middle word missing, as indicated by a blank line. The instructions were as follows: “You are going to see steps (in a recipe or how to perform a managing finance task) one at a time. However, 1 word in each step will be missing. Fill in the blank at each step with the most logical choice. Please read the steps aloud. Try to remember the steps because I will ask you about them later.” For example, if a participant was asked to prepare an omelet, 1 step of the task was presented as follows: “beat together _____” and the participant could generate the appropriate item to complete the step (eg, egg or eggs).

Generation effect protocol: design and procedure 

Figure 1 provides an illustration of the study design and procedural flow of the generation effect protocol. One meal preparation task and 1 managing finance task were carried out in the provided condition and 1 meal preparation task and 1 managing finance task were carried out in the generated condition. Tasks were counterbalanced for condition (generated vs provided) and IADL task category (meal preparation vs managing finances). Thus, if meal preparation tasks were presented first for 1 participant, then managing finance tasks were presented first for the second participant. Within the task categories, if the French toast task was presented first in a generated condition for 1 participant, the second participant was presented with the French toast task in a provided condition after learning the omelet task in a generated presentation. Thus, participants first learned 2 tasks that could be either related to managing finances or meal preparation. Task presentations were followed by immediate recall, in which participants were asked to verbally recall the task steps required to perform each of them. Thirty minutes after initial presentation of the task, participants were then required to actually perform the tasks. During the 30 minutes between initial presentation and the performance recall of the task, participants were asked to complete neuropsychologic tests. After this, participants learned the 2 other tasks in a similar way to the procedure described above. Finally, 1 week after initial presentations, participants were contacted by phone to assess verbal recall of task steps presented for all 4 tasks.

  • View full-size image.
  • Fig 1. 

    Study design and procedural flow of the generation effect protocol in both MS and control groups. Note that the order of tasks and conditions was counterbalanced for each participant. Abbreviation: NP, neuropsychologic testing.

Neuropsychologic testing 

MS can cause cognitive impairments in the areas of attention, information processing speed,33, 34 learning and memory,6, 7, 8 executive functions,35, 36 and visuospatial abilities and all of these abilities could potentially impact the degree of benefit one could gain from a self-generation paradigm. Therefore, the battery of neuropsychologic tests administered was geared toward assessing these cognitive functions.

Neuropsychologic testing: attention/concentration and processing speed 

In the WAIS-R digit span forward trial,37 the participant was instructed to repeat the string of digits in the same order in which they were presented by the examiner. In the digit span backward trial, the participant was instructed to repeat the string of digits in the reverse order from which they were presented. The dependent measure was the number of the correct responses for each trial.

Neuropsychologic testing: SDMT oral version 

The SDMT38 involved a set of 9 meaningless geometric figures, each of which corresponds to a different number (1–9). A key was presented at the top of the page, indicating which number corresponded to each symbol. Participants were then presented with the geometric figures only and asked to quickly state the number associated with each figure (with the key in their sight for reference). The dependent measure was the total number of correctly completed numbers in 90 seconds. The SDMT has been found to be a valid and responsive measure of processing speed in persons with MS.39, 40, 41, 42, 43

Neuropsychologic testing: executive control 

We used 3 selected subtests of the D-KEFS.44 In the TMT, the score from the number-letter switching task was used as the dependent variable. This task is designed to assess flexibility of thinking on a visuomotor sequencing task. Time to completion (in seconds) was used as the dependent variable and reported in table 1.

The second subtest, the Verbal Fluency Test, assesses fluent productivity in the verbal domain. The task required the verbal generation of words in 60 seconds to the letter prompts F, A, and S. The number of words generated with the 3 letters was summed and used as the dependent variable.

The Color-Word Interference Test task assesses susceptibility to interference and the ability to shift perceptual set as the examinee must inhibit an over-learned verbal skill (reading) to perform a color-naming task. Time to completion on the shifting/inhibition and switching was used as the dependent variable. Time was recorded in seconds.

In addition, for all 3 D-KEFS subtests, raw scores were converted to scaled scores with a mean ± SD of 10±3, corrected for age.

Neuropsychologic testing: language functions 

The BNT45 consists of 60 line drawings of common objects for confrontational naming. The BNT has been widely used in the assessment of patients with aphasia, dementia, and stroke and is effective in finding naming impairments in patients.46 The dependent variable was the total number correct.

Neuropsychologic testing: verbal memory 

The CVLT47 consists of a list of 16 words from 4 semantic categories. Each of the 16 items in each CVLT list belongs to 1 of 4 categories of shopping items. This list (list A) was presented over 5 trials, and was immediately followed by a second list (list B) of 16 words presented for only 1 trial. Delayed recall and recognition of list A items followed the immediate recall of list B. Total correct for all 5 trials of list A was recorded. Recognition was assessed by performance on the target and/or distractor items at the 20 minutes delayed recall. The CVLT discriminability and response bias were calculated.

Neuropsychologic testing: emotional functioning 

The STAI48 is a standardized, well-established measure of anxiety. Higher STAI state and trait scores represent more anxiety. Both STAI state and trait scores were used in analysis.

The CMDI49 is a 42-item self-report measure designed to assess depression in MS and other medical populations. The overall CMDI score was used in the present study as a measure of depressive symptomatology.

Procedure 

We recruited participants by advertisements distributed at local support groups and clinics. On initial phone contact, potential participants were screened for participation based on the inclusion and exclusion criteria discussed above. Participants were then scheduled for interview and testing. All recruitment and experimental procedures were approved by the institutional review board and Health Insurance Portability and Accountability Act compliance boards. Before study enrollment, all participants indicated willingness to participate in the study by signing a consent form approved by the institutional review board.

The testing session lasted approximately 3 hours, in which the neuropsychologic battery and the self-generation protocol were administered. One week after testing participants were contacted by telephone to test verbal recall of the stimuli presented within the self-generation protocol. All participants were paid for their participation.

Data Analysis 

We analyzed cognitive and emotional functioning using descriptive statistics and 1-way ANOVA to compare performance between MS and healthy control groups.

Performance on the generation effect protocol was analyzed by a 2 (group) by 2 (condition) by 2 (task) by 3 (time), mixed-design repeated-measures ANOVA, with group as the between group factor and condition, type of task and trial as within-group factors. To evaluate the effect size in the above analyses, η2 was used.50 When needed, we used contrast tests to examine repeated effects.

Back to Article Outline

Results 

Neuropsychologic and Emotional Functioning 

Cognitive and emotional functioning for the MS and healthy control groups are presented in table 2. In terms of cognitive functioning, participants with MS performed similarly to control participants on all tasks except for the D-KEFS. Specifically, the MS group took significantly longer (mean scaled score, 8.1) to perform the D-KEFS TMT: number letter switching task than the control group (mean scaled score, 11.1) (F1,37=10.8, P=.002; η2=.23). On the D-KEFS Verbal Fluency Test letter fluency subtest, the MS group generated significantly fewer words (mean scaled score, 8.7) than healthy control participants (mean scaled score, 12.1) (F1,37=7.9, P=.008; η2=.18). Finally, on the D-KEFS Color-Word Interference Test: inhibition and switching task, the MS group required significantly more time to completion (mean scaled score, 9.1) than the control group (mean scaled score, 11.5) (F1,37=4.3, P=.04; η2=.11). With regard to emotional functioning, scores on the CMDI and STAI did not differ significantly between groups, and fell within the normative range when compared with published normative data.

Table 2. Neuropsychologic Test Performance
Domain AssessedMS (n=20)Healthy Controls (n=18)Fη2
Speed of processing
SDMT47.3±12.946.2±14.30.06.002
Episodic memory
CVLT sum of 5 trials54.6±10.658.7±8.11.7.04
CVLT discriminability94.2±7.594.6±5.00.05.002
Working memory
Digit span forward9.7±2.611.0±2.03.1.08
Digit span backward7.5±2.78.3±2.90.83.02
Executive functions
D-KEFS TMT: number-letter switching105.7±44.168.5±19.610.8†.23
D-KEFS Verbal Fluency Test letter fluency33.5±13.145.1±12.17.9†.18
D-KEFS Color-Word Interference Test: inhibition and switching73.1±29.956.6±15.14.3⁎.11
Toglia category assessment30.3±3.330.7±2.70.18.005
Deductive reasoning test18.8±1.619.6±1.32.8.07
Language
BNT54.9±4.151.7±8.02.4.06
Affect symptomatology
STAI state anxiety35.1±13.230.6±7.91.6.06
STAI trait anxiety38.8±11.738.7±8.90.001.001
CMDI80.3±27.071.3±19.81.2.03

NOTE. Values are mean ± SD or as otherwise indicated.

⁎P<.05;

†P<.01.

Generation Effect 

Generated versus provided conditions 

Performance on the generation effect protocol is presented in figure 2. Tasks (ie, managing finances, meal preparation) that were learned under the generated condition were recalled at a significantly higher rate (mean, 16.8) than tasks that were presented in a provided condition (mean, 14.7), collapsed across group and time. The main effect of condition (generated vs provided) was significant, and showed a medium effect size (F1,36=28.1, P<.001; η2=.43), showing the significant benefit of self-generation over provided presentations across the tasks learned.

  • View full-size image.
  • Fig 2. 

    Recall scores (mean and SE) for generated versus provided conditions at each time period across tasks and groups. Time: immediate verbal recall after initial presentation; performance recall 30 minutes after initial presentation; 1-week verbal recall after initial presentation.

MS group versus control group 

The MS and control groups did not differ significantly in the mean number of task items recalled collapsed across time and condition (F1,36=.40, P=.50; η2=.01). Additionally, the relative difference between the provided and generated conditions was equivalent across participant groups: the interaction of participant group (control vs MS) and condition (generated vs provided) was not significant and showed a small effect size (F1,36=.25, P=.61; η2=.007). Thus, both control and MS groups showed a similar pattern: both groups benefited from the generated condition, compared with the provided condition, when learning the meal preparation and managing finances tasks.

Time 

As expected, the recall of task items dropped significantly across time in both the generated and provided conditions (F2,35=55, P<.001; η2=.75). Contrast tests revealed a significant difference between recall 30 minutes and 1 week after task presentation (F1,35=11.3, P<.001; η2=.75). However, the main effect of time was moderated by a significant 3-way interaction between group membership, learning condition, and time of recall (F2,35=3.4, P=.04; η2=.16). Contrast tests to examine repeated effects revealed a significant difference between recall immediately and 30 minutes after task presentation (F1,36=5.6, P=.02; η2=.13), but not between recall 30 minutes after task presentation and recall after 1 week. Participants in each group (control, MS) showed a distinct pattern of performance across the 3 time periods and the 2 learning conditions (fig 3, table 3). To examine the specific pattern of group performance, 2 (task) by 2 (condition) by 3 (time) repeated-measures ANOVA was done separately for each of the groups. The control group showed the greatest benefit from self-generation at immediate recall (F1,17=11.4, P=.004; η2=0.4). The MS group showed the greatest benefit from self-generation at the 1-week recall, as seen in figure 3 (F1,19=3.8, P=.06; η2=.17).

  • View full-size image.
  • Fig 3. 

    Recall scores for generated versus provided tasks at each time period for the MS participants and for the healthy controls. Time: immediate verbal recall after initial presentation; performance recall 30 minutes after initial presentation; 1-week verbal recall after initial presentation. Abbreviation: HC, healthy controls.

Table 3. Means for the Generated and Provided Conditions Across Times and Tasks for the Controls and MS Participants
ConditionsMS Group (n=20)Control Group (n=18)
Generated ConditionProvided ConditionGenerated ConditionProvided Condition
Immediate recall18.0±0.7516.3±0.6818.1±0.7914.9±0.87
Performance recall (30min)18.9±0.5317.0±0.6317.0±0.6716.2±0.78
One-week recall14.5±0.9111.5±0.8014.1±0.7712.6±0.89

NOTE. Values are mean ± SE.

Task (managing finances versus meal preparation) 

The main effect of task type was not significant (F1,36=.07, P=.79; η2=.002), indicating that the meal preparation and managing finances tasks did not differ in the mean number of items recalled overall. The interaction of type of task and condition (ie, generated vs provided) was not significant (F1,36=.40, P=.07; η2=.09) (fig 4). To compare the relative effect of self-generation on each of the 2 types of task, we computed a learning retention score by subtracting the score obtained in the provided condition in each task from the score obtained in the generated condition of each task. A paired t test was performed to compare the learning retention scores that were obtained in the managing finances and the meal preparation across the 2 participant groups. Results showed a significant difference between the learning retention scores, where self-generation had a stronger effect on the meal preparation task than on the managing finances task for all participants (t=1.9; 2-tailed, P=.05).

Post hoc (order effect) 

Because tasks and learning conditions were counterbalanced, a 2 (order) by 2 (task) by 2 (condition) by 3 (time) repeated-measures ANOVA was performed to examine potential of order effects. The analysis did not reveal any significant order effect (F1,44=.51, P=.47).

Back to Article Outline

Discussion 

Results of the present study show that self-generation enhances new learning and memory for functional tasks related to IADLs in both persons who are healthy and persons with MS. The results obtained in this study extend previous research on self-generation12, 13, 25, 28, 51, 52 and extends the known benefits of self-generation to everyday life activities, and is consistent with another study that has recently addressed this issue.29

Self-generation strategy significantly improved learning and memory for all functional tasks. It was interesting to note, however, that this effect was stronger in the meal preparation tasks than in the managing finances tasks. This finding could potentially be explained by the role of task meaningfulness. Research evidence strongly suggests that functional performance is enhanced when participants perceive the activity as having meaning and purpose.53 A contributing factor in whether activities are perceived as meaningful and purposeful is their contextual relevancy, as well as the goal of the task. The meal preparation tasks could be more meaningful than managing finance tasks utilized in this study for a few reasons. First, the participants performed the actual tasks in both managing finance and meal preparation; however, they used actual stimuli in the meal preparation task, but mock stimuli in the managing finances tasks (eg, paying a mock electric bill with a mock check). Second, the meal preparation task provided a potential incentive to the participants by allowing them to eat the food they cooked once the task was completed. There was no explicit reward to performing the managing finance task successfully. These explanations are also supported by Ma,54 Marteniuk,55 Bloch,56 and colleagues who found that a person's actions are guided by perceived information, such as, what is the goal of the task. These various explanations are not mutually exclusive and may have interacting roles and influence in obtaining the observed effect. Thus, task conditions (eg, meaningfulness, familiarity, complexity) are likely important factors, because they either limit or facilitate the ability of self-generation to improve functional, everyday-life memory. When considering assigning a functional task in rehabilitation, it is important to take into consideration the fact that each task may have contextual and conceptual characteristics that can interact with the learner's characteristics and may affect their effectiveness in rehabilitation.

The current study documented the continued benefit of learning through self-generation after a 1-week delay when everyday life tasks are examined. It is important to note that in the Chiaravalloti and DeLuca24 laboratory-based study, after a 1-week delay, the benefit of self-generation was lost in both the healthy control and MS groups.

One possible explanation for the difference found in 1-week recall performance between previous laboratory-based studies and the current functional study could also be related to the meaningfulness of tasks to be learned. It is well known that recall is better for information that is more personally meaningful or salient.57, 58, 59 Some support for this explanation can be found in Buyer and Dominowski's study,60 where the benefit of self-generation after a 1-week delay in healthy university students using a lab-based task was seen on only the more challenging items. These authors examined how degree of cognitive effort can help enhance the generation effect in laboratory tasks. They found that retention after a 1-week delay increased as the cognitive effort required generating a phrase increased. When participants were required to generate easy items, the degree of benefit from self-generation diminished after 1 week. They concluded that representation in semantic memory as a functional unit is a primary factor for the emergence of the generation effect. Another potential explanation for the difference in the generation effect results between lab-based tasks and functional tasks at 1-week recall also can be based on the multifactor theory of the generation effect. The multifactor theory61 postulates that in addition to increasing attention to and processing of each encoded word (item-specific processing11), the process of self-generation also increases the frequency of encoding inter-relations between the target and other aspects of the study session (relational processing61). Functional tasks provide inter-relations between the environment present during learning as well as the actual performance of the task. Due to this increase in encoded inter-relations, one could expect self-generation to be more beneficial in functional tasks than in lab-based tasks.

Study Limitations 

Although the current study supports the efficacy of self-generation to improve memory for everyday tasks, it does have some important limitations. First, this research was conducted with a relatively small number of participants with MS who do not represent the full extent of the course of MS, particularly with respect to degree of cognitive impairment. Furthermore, participants in this study did not have important memory impairments. Level of cognitive impairment has been shown to influence the degree to which a given person can benefit from cognitive rehabilitation in previous research.62 To date, we have found 2 studies that compared the degree of benefits of self-generation in persons with MS with different levels of cognitive impairment using lab tasks.16, 29 The findings of these 2 studies showed that despite significant memory impairment in the impaired MS group, self-generated encoding enhanced memory to the same extent as the control group and the MS group without cognitive impairments.

The second study limitation relates to the study procedures. The results of this study could have been more robust if participants had been randomized when assigned to the learning conditions, rather than alternate assignment. Future studies should consider this procedure.

The nature of the tasks used in the present study represent a third study limitation. The present study used 2 meal preparation tasks (prepare an omelet and French toast) and 2 managing finances tasks (buying clothes from a catalog, paying an electric bill). These 2 tasks were chosen for 2 reasons: (1) tasks in each domain of functioning needed to be equated for level of difficulty (2) these tasks are IADL tasks, and most people need to be capable of performing IADLs to a certain level to function independently. However, the tasks assigned may have been over learned, and likely include procedural memory elements in their successful completion. Despite this, a clear benefit of self-generation was still observed. Future studies examining the effect of self-generation should use meaningful functional tasks that are less familiar to participants. Last, the fact that participants actually performed the tasks 30 minutes earlier (as opposed to simply recalling the items), may have led to increased encoding and improved recall 1 week later. However, although this enhanced encoding may have had beneficial effect at 1 week across both the provided and generated conditions, it clearly was significantly more beneficial when items were self-generated. As such, it is important for future studies to explore the factors that are likely to maximize the self-generation effect including (1) the contextual characteristics of the tasks (eg, meaningful types of tasks), (2) the actual performance of the tasks, and (3) what MS subpopulations (eg, degree of cognitive impairment) will benefit the most from the self-generation strategy.

Back to Article Outline

Conclusions 

The ultimate goal of the present study was to examine whether increased encoding associated with self-generation would improve recall and performance of everyday life functional tasks. By improving the ability of persons with MS to learn new tasks, we hope to improve the quality of their lives at work, home, or in social interactions. The results of the present study suggest that self-generation may hold substantial promise for improving the lives of persons with MS showing new learning and memory deficits.

Back to Article Outline

References 

  1. Rao SM, Grafman J, DiGiulio DV, et al. Memory dysfunction in multiple sclerosis: its relation to working memory, semantic encoding, and implicit learning. Neuropsychology. 1993;7:364–374
  2. Bobholz JA, Rao SM. Cognitive dysfunction in multiple sclerosis: a review of recent developments. Curr Opin Neurol. 2003;16:283–288
  3. Beatty WW, Blanco CR, Wilbanks SL, Paul RH, Hames KA. Demographic, clinical, and cognitive characteristics of multiple sclerosis patients who continue to work. J Neurol Rehabil. 1995;9:167–173
  4. Rao SM, Leo GJ, Ellington L, Nauertz T, Bernardin L, Unverzagt F. Cognitive dysfunction in multiple sclerosis (II. Impact on employment and social functioning). Neurology. 1991;41:692–696
  5. Kessler HR, Cohen RA, Lauer K. The relationship between disability and memory dysfunction. J Neurosci. 1992;62:17–34
  6. DeLuca J, Barbieri-Berger S, Johnson SK. The nature of memory impairments in multiple sclerosis: acquisition versus retrieval. J Clin Exp Neuropsychol. 1994;16:183–189
  7. DeLuca J, Gaudino EA, Diamond BJ, Christodoulou C, Engel RA. Acquisition and storage deficits in multiple sclerosis. J Clin Exp Neuropsychol. 1998;20:376–390
  8. Gaudino EA, Chiaravalloti ND, DeLuca J, Diamond BJ. A comparison of memory performance in relapsing-remitting, primary progressive and secondary progressive, multiple sclerosis. Neuropsychiatry Neuropsychol Behav Neurol. 2001;14:32–44
  9. Demaree HA, Gaudino EA, DeLuca J, Ricker JH. Learning impairment is associated with recall ability in multiple sclerosis. J Clin Exp Neuropsychol. 2000;22:865–873
  10. Thornton AE, Raz N. Memory impairment in multiple sclerosis: a quantitative review. Neuropsychology. 1997;11:357–366
  11. Begg I, Snider A, Foley F, Goddard R. The generation effect is no artifact: generating makes words distinctive. J Exp Psychol Learn Mem Cogn. 1989;15:977–989
  12. Marsh E, Edelman G, Bower GH. Demonstrations of a generation effect in context memory. Mem Cognit. 2001;29:798–805
  13. Slamecka NJ, Graf P. The generation effect: delineation of a phenomenon. J Exp Psychol [Hum Learn]. 1978;4:592–604
  14. Slamecka NJ, Fevreiski J. The generation effect when generation effect fails. J Mem Lang. 1983;22:153–163
  15. Craik FI, Lockhart RS. Levels of processing: a frame work for memory research. J Mem Lang. 1972;11:671–684
  16. O'Brien A, Chiaravalloti N, Arango-Lasprilla JC, Lengenfelder J, DeLuca J. An investigation of the differential effect of self-generation to improve learning and memory in multiple sclerosis and traumatic brain injury. Neuropsychol Rehabil. 2007;17:273–292
  17. Lengenfelder J, Chiaravalloti ND, DeLuca J. The efficacy of the generation effect to improve new learning in persons with traumatic brain injury. Rehabil Psychol. 2007;52:290–296
  18. Smith ML. Recall of frequency of occurrence of self-generated and examiner-provided words after frontal or temporal lobectomy. Neuropsychologia. 1996;34:553–563
  19. Hsieh S, Lee C. Source memory in Parkinson's disease. Percept Mot Skills. 1999;89:355–367
  20. Barrett AM, Crucian GP, Schwartz RL, Heilman KM. Testing memory for self-generated items in dementia: method makes a difference. Neurology. 2000;54:1258–1264
  21. Dick MB, Kean ML, Sands D. Memory for internally generated words in Alzheimer-type dementia: breakdown in encoding and semantic memory. Brain Cogn. 1989;9:88–108
  22. Mitchell DB, Hunt RR, Schmitt FA. The generation effect and reality monitoring: evidence from dementia and normal aging. J Gerontol. 1986;41:79–84
  23. Multhaup KS, Balota DA. Generation effects and source memory in healthy older adults and in adults with dementia of the Alzheimer type. Neuropsychology. 1997;11:382–391
  24. Chiaravalloti ND, DeLuca J. Self-generation as a means of maximizing learning in multiple sclerosis: an application of the generation effect. Arch Phys Med Rehabil. 2002;83:1070–1079
  25. Crutcher RJ, Healy AF. Cognitive operations and the generation effect. J Exp Psychol Learn Mem Cogn. 1989;15:669–675
  26. Gardiner JM, Rowley JM. A generation effect with numbers rather than words. Mem Cognit. 1984;12:443–445
  27. Jurica PJ, Shimamura AP. Monitoring item and source information: evidence for a negative generation effect in source memory. Mem Cognit. 1999;27:648–656
  28. Kinjo H, Snodgrass JG. Does generation effect occur for pictures?. Am J Psychol. 2000;113:95–121
  29. Basso MR, Lowery N, Ghormley C, Combs D, Johnson J. Self-generated learning in people with multiple sclerosis. J Int Neuropsychol Soc. 2006;12:640–648
  30. Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol. 1983;13:227–231
  31. Wilkinson GS. Wide Range Achievement Test−3: administration manual. Wilmington: Wide Range; 1993;
  32. Fisher AG. Assessment of motor and process skills. 4th ed. Fort Collins: Three Star Pr; 2001;
  33. Schulz D, Kopp B, Kunkel A, Faiss JH. Cognition in the early stage of multiple sclerosis. J Neurol. 2006;253:1002–1010
  34. Goverover Y, Genova HM, Hillary GM, DeLuca J. The relationship between neuropsychological measures and the timed instrumental activities of daily living task in multiple sclerosis. Mult Scler. 2007;13:636–644
  35. Pozzilli C, Bastianello S, Padovani A, et al. Anterior corpus callosum atrophy and verbal fluency in multiple sclerosis. Cortex. 1991;27:441–445
  36. Arnett PA, Rao SM, Grafman J, et al. Executive functions in multiple sclerosis: an analysis of temporal ordering, semantic encoding, and planning abilities. Neuropsychology. 1997;11:535–544
  37. Wechsler D. Wechsler Adult Intelligence Scale: revised manual. New York: Harcourt Brace Jovanovich; 1981;
  38. Smith A. Symbol Digits Modalities Test. Los Angeles: Western Psychological Services; 1982;
  39. Beatty WW, Goodkin DE, Monson N, Beatty PA. Anterograde and retrograde amnesia in patients with chronic progressive multiple sclerosis. Arch Neurol. 1988;45:611–619
  40. Camp SJ, Stevenson VL, Thompson AJ, et al. Cognitive function in primary progressive and transitional progressive multiple sclerosis: a controlled study with MRI correlates. Brain. 1999;122:1341–1348
  41. Franklin GM, Heaton RK, Nelson LM, Filley CM, Seibert C. Correlation of neuropsychological and MRI findings in chronic/progressive multiple sclerosis. Neurology. 1988;38:1826–1829
  42. Sperling RA, Guttmann CR, Hohol MJ, et al. Regional magnetic resonance imaging lesion burden and cognitive function in multiple sclerosis: a longitudinal study. Arch Neurol. 2001;58:115–121
  43. Swirsky-Sacchetti T, Mitchell DR, Seward J, Gonzales C. Neuropsychological and structural brain lesions in multiple sclerosis: a regional analysis. Neurology. 1992;42:1291–1295
  44. Delis DC, Kaplan E, Kramer JH. Delis-Kaplan executive function system. San Antonio: Psychological Corp; 2000;
  45. Kaplan E, Goodglass H, Weintraub S. The Boston Naming Test. 2nd ed. Philadelphia: Lea & Febiger; 1983;
  46. Lezak MD. Neuropsychological assessment. 3rd ed. New York: Oxford Univ Pr; 1995;
  47. Delis DC, Kramer JH, Kaplan E, Ober BA. California Verbal Learning Test: research edition (Version I). San Antonio: Psychological Corp; 1987;
  48. Spielberger C. Manual for the State-Trait Anxiety Inventory. Palo Alto: Mind Garden Publishers; 1983;
  49. Nyenhuis DL, Luchetta T, Yamamoto C, et al. The development, standardization, and initial validation of the Chicago Multiscale Depression Inventory. J Pers Assess. 1998;70:386–401
  50. SPSS Inc. SPSS base applications guide. Chicago: SPSS Inc; 1997;
  51. Mulligan NW, Lozito JP, Rosner ZA. Generation and context memory. J Exp Psychol Learn Mem Cogn. 2006;32:836–846
  52. Johnson MK, Raye CL, Foley HJ, Foley MA. Cognitive operations and decision bias in reality monitoring. Am J Psychol. 1981;94:37–64
  53. Bloch MW, Smith DA, Nelson DL. Heart rate, activity, duration, and affect in added-purpose versus single purpose jumping activities. Am J Occup Ther. 1989;43:25–30
  54. Ma H, Trombly CA, Robinson-Podolski C. The effect of context on skill acquisition and transfer. Am J Occup Ther. 1999;53:138–144
  55. Marteniuk RG, MacKenzie CL, Jeannerod M, Athenes S, Dugas C. Constraints on human arm movement trajectories. Can J Psychol. 1987;41:365–378
  56. Bloch MW, Smith DA, Nelson DL. Heart rate, activity, duration, and affect in added-purpose versus single purpose jumping activities. Am J Occup Ther. 1989;43:25–30
  57. Bower G, Gilligan S. Remembering information related to one's self. J Res Pers. 1979;13:420–432
  58. Craik FI, Tulving E. Depth of processing and the retention of words in episodic memory. J Exp Psychol Gen. 1975;104:268–294
  59. Westmacott R, Moscovitch M. The contribution of autobiographical significance to semantic memory. Mem Cognit. 2003;31:761–774
  60. Buyer LS, Dominowski RL. Retention of solutions: it is better to give than to receive. Am J Psychol. 1989;102:353–363
  61. Marsh EJ, Edelman G, Bower GH. Demonstrations of a generation effect in context memory. Mem Cognit. 2001;29:798–805
  62. Chiaravalloti ND, DeLuca J, Moore NB, Ricker JH. Treating learning impairments improves memory performance in multiple sclerosis: a randomized clinical trial. Mult Scler. 2005;11:58–68

 Published online June 30, 2008 at www.archives-pmr.org.Supported by the National Multiple Sclerosis Society (grant no. RG3837A1/T) and the Henry H. Kessler Foundation.No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.

PII: S0003-9993(08)00307-9

doi:10.1016/j.apmr.2007.11.059

Archives of Physical Medicine and Rehabilitation
Volume 89, Issue 8 , Pages 1514-1521, August 2008

Article Tools

* Image Usage & Resolution