|Running Head: MEMORY FOR ACTIONS AND REPETITIONS|
Does Combining SPTs and VTs Increase the Memorability Of Presented Material?Jeff Brunsgaard, Tina Knudson, Carmen May, Amy Ugalde, and Gloria Marmolejo Winona State University - Rochester AbstractWe investigated the influence of deep actions on memory for massed retrieval repetition. The actions were subject-performed tasks and verbal tasks using pictures. Twenty-one subjects viewed 30 different pictures, performing a verbal task, (name), or a subject-performed task along with a verbal task (name and write or name and sketch). Of the 30 pictures, six were circled and participants were instructed to remember all six with their actions. Following a five-minute filler task, participants attempted to retrieve the six mandatory pictures plus four additional ones and perform their corresponding actions consecutively through six trials. Finally, participants took a test to measure picture, action, and repetition discrimination. Results showed that picture recall was affected by the action, and memory increased with retrieval repetitions. Name and write tasks were often less recalled than name and sketch tasks over retrieval repetitions. Action recognition was best for name and sketched pictures and significantly lower for named pictures. The data suggests that the specific actions and retrieval repetition moderate the subject-performed task effect. Does Combining SPTs and VTs Increase the MemorabilityOf Presented Material? Does a combination of tasks performed with material that is to be learned increase the memory of such material, as the depth of processing approach to memory (Matlin, 2002) would indicate? Would the effects of such a combined task hold true consistently over repeated recall trials? Previous research has looked into the effect of subject-performed tasks (hereafter SPTs) on the recall of presented material over multiple trials, and has also looked into the effect of verbal tasks (hereafter VTs) on the same type of recall (Marmolejo & Krizan, 2001). In the past, SPTs and VTs had been separated into two different categories of action (Cohen, 1981), but more recent research conducted by Marmolejo & Krizan (2001) indicates that perhaps SPTs and VTs are more closely related than was previously thought to be the case. Namely, this research showed that rather than being two separate categories of action, SPTs and VTs might simply be two disparate ends of a continuum of action that influences memory. Engelkamp and Dehn (2000) have indicated that, overall, SPTs improve item encoding to a greater extent than do experimenter-performed tasks (EPTs). In this case the SPT effect was attributed to the uniqueness in memory of items that are associated with tasks that require the activation of a greater portion of the subject¡¦s neural network in order to perform a task. This stands in contrast to EPTs, which are simply presented to the subject and do not require any additional neural activation and are therefore less memorable. Hypotheses for Increased Memory due to the SPT Effect There are numerous possible reasons for the increase in memorability that is characterized by the SPT effect which, due to the current research methodology, cannot be distinguished from one another. The two that seem most applicable here are the richness of encoding and non-strategic processing that is inherent to SPTs. Richness of encoding would indicate that the SPT effect is due to the fact that SPTs provide multiple avenues for information to be remembered, with such characteristics as tactile sensation, movement, and other modes of perception being incorporated into the memory for the item that is being encoded, rather than just the visual or verbal stimuli provided by an EPT or simple presentation of information. Also worth consideration in a discussion of richness of encoding is the generation effect, which posits that tasks that require the subject to generate information themselves are more memorable than tasks in which subjects simply receive the information without doing anything to manipulate it themselves. Secondly, non-strategic processing encourages incidental learning of targets associated with SPTs that involve more complex tasks such as sketching or enacting something, whereas simpler SPTs involving mainly linguistic tasks such as speaking a name aloud encourage more strategic processing, which may seem to increase their memorability on testing. This means that subjects are more likely to use metamemory strategies such as rehearsal and association on linguistic tasks, while their memory for motor tasks is a function of the fact that these tasks cause more stimulation. This seeming increase in memory for verbal tasks on some previous tests is not necessarily the case, since most forms of testing currently in use test for the more strategic aspects of processing by asking subjects to either write or recognize the items (associating test items with the means of encoding), as opposed to the multi-modal approach that the more complex SPTs are encoded with and encourage memorability with. To help negate this effect of the style of testing, we tested the subjects by having them recall the learned items in the same way in which they encoded them. Often times hypermnesia, a net gain in recalled items, is obtained over repeated recall trials involving SPTs, as shown by Olofsson¡¦s 1997 study. This effect is also shown for the general recall of pictures and words independent of any actions that may or may not have been performed with them. Therefore, in an experiment that tests a combination of these effects over repeated free and mandatory recall trials, we would expect that an effect of hypermnesia would be demonstrated, despite Krizan and Marmolejo¡¦s 2001 conclusions to the opposite effect. The combinations of SPTs that we have decided to test here emphasize a combination of two tasks that are frequently performed in conjunction with each other in everyday life; namely, people often perform a simple linguistic task, such as reminding themselves verbally of tasks that they have yet to perform, or the individual steps in an ongoing task, as they are performing a more complex task at the same time. This combination of tasks could well make the events that they are performed in conjunction with more memorable, and therefore easier to learn in the present and recall for future use. The fact that multiple sources of input provide a more meaningful experience of the event would encourage its placement into episodic memory, as opposed to simply forgetting it. This would tend to indicate that SPTs will combat the hypomnesia effect, in which subjects show net item losses over multiple recall trials. The placement in episodic memory which would be encouraged by SPTs would put them into a more lasting form of memory than the working memory, in which item losses are common, and thereby make the memory for that event more lasting. Purpose The current experiment had one primary goal, which was to test the effect of the richness of the encoding condition on subject¡¦s recall over multiple memory trials. Realizing that the distinction between SPTs and VTs is somewhat arbitrary, we had subjects perform a combination of tasks in conjunction with the presented stimulus. The tasks performed include naming the stimulus aloud (poorest encoding), naming the stimulus and writing the name (median encoding), and naming the stimulus while sketching it (richest encoding). Our hypothesis indicated that if richness of encoding influences the SPT effect, we would expect to see a significantly better rate of recall and recognition for items that had richer information present at encoding than did those items that had less information or stimulation present at encoding. On the other hand, if the SPT effect is due to a process other than richness of encoding, we would expect to see no difference in recall between the various encoding conditions. We also expected that we would obtain a hypermnesic effect over repeated conditions, mostly because this is the result that has been obtained by the largest body of previous research into free recall. A hypermnesic effect was expected to occur across all conditions, as this appears to be due to the process of reinforcement of information over repeated trials, rather than an aspect of the encoding conditions. AbstractWe investigated the influence of deep actions on memory for massed retrieval repetition. The actions were subject-performed tasks and verbal tasks using pictures. Twenty-one subjects viewed 30 different pictures, performing a verbal task, (name) or a subject-performed task along with a verbal task (name and write or name and sketch). Of the 30 pictures, six were circled and participants were instructed to remember all six with their actions. Following a five-minute filler task, participants attempted to retrieve the six mandatory pictures plus four additional ones and perform their corresponding actions consecutively through six trials. Finally, participants took a test to measure picture, action, and repetition discrimination. Results showed that picture recall was affected by the action, and memory increased with retrieval repetitions. Name and write tasks were often less recalled than name and sketch tasks over retrieval repetitions. Action recognition was best for name and sketched pictures and significantly lower for named pictures. The data suggests that the specific actions and retrieval repetition moderate the subject-performed task effect.
MethodParticipantsTwenty-one students who volunteered were recruited from the University Center of Rochester, MN. Most of them were enrolled in a psychology course and given one extra credit point for each quarter hour involved in the experiment. Their age ranged between 18-51 years old. Design and Materials The study was a 3 x 6 two-way within-subjects design. Three levels of action, naming, naming and writing, and naming and sketching were combined with six levels of Recall Repetition, R1 to R6. A slide show presented thirty pictures of concrete objects from Snodgrass and Vanderwart¡¦s (1980) norms. These pictures were of animals, possible household furniture, fruit and vegetables, weapons, birds, and vehicles. There were three PowerPoint presentations to be displayed in Compaq and Hewlett Packard 90 MHz Pentium computers with 14-inch monitors. Three sets were formed with the same 30 pictures in a different random order. Each picture was shown on the screen for five seconds and followed by a black ¡§action request¡¨ screen. The action requested was name, name and write, or name and sketch in white print. This screen allowed 20 seconds for the participant to ¡§practice¡¨ their action request. All actions were counterbalanced across the three sets, so that all three actions were performed with every picture across participants. Six of the 30 pictures in each set were enclosed within a red circle, giving a total of 18 different circled pictures across three picture sets. Actions requested were also counterbalanced across the three sets for the same reason as above. Two pictures were to be named, two named and written, and two named and sketched. Participants¡¦ actions of writing or sketching were put on a piece of blank white paper for each item requesting name and write or name and sketch. Naming responses were given out loud for the experimenter to observe. The five-minute filler task was an internet crossword puzzle. The test that examined recognition of the pictures, respective actions and repetitions contained the names of 36 items randomly listed. Six of them were not previously showed in the presentation and six of the 30 listed were the mandatory circled pictures. Three columns were reserved for responses regarding picture recognition, action performed, and specific repetition(s). After being randomly assigned to one of the three picture sets, each participant was instructed to pay close attention to the slides and perform the appropriate action for every picture according to the ¡§action request¡¨ screen. For example, when presented with the picture of a rocking chair, followed by the action request screen ¡§name and write¡¨, the participant named the rocking chair out loud and then wrote rocking chair on a blank piece of paper. If the instruction was to name and sketch, the participant would name it out loud and then draw it on a piece of blank paper. If the instruction was to just name the object, the participant would simply say what it is out loud. This was done for all 30 pictures. Participants were informed that some pictures would appear inside a red circle, for which they should circle the picture along with the action to be performed with it. They received three practice trials before the 30 experimental trials. After working on the filler task for five minutes, participants proceeded to recall all six circled mandatory pictures along with their respective actions for each six trials given. Also, the participants were asked to recall four non-circled pictures with the appropriate actions, and for each trial to follow, one additional non-circled picture with the appropriate action(s). In the last stage, participants completed the recognition test. For each item, subjects answered the question, ¡§Saw it in the slide show?¡¨ by recording a yes/no response. If the answer was ¡§yes¡¦, they proceeded to answer ¡§Action performed?¡¨ with a response of name, name and write, name and sketch, or none, followed by answering ¡§Performed in trials?¡¨ with the list of the trial number(s) 1-6 in which the action for the specific picture was performed. The experiment took between 40 to 60 minutes. Results and Discussion
This study examined whether the different combinations of SPTs and VTs and the involvement of motor programs, and multi-modal representations, could influence the SPT effect over 6 trial repetitions. The question of whether massive repetitive retrieval influenced the SPT effect and interference with what actions were performed with what picture and when, was also explored. Recall and recognition data were analyzed by using ANOVAs with Ñ=.05. Only the 6 mandatory pictures were reported in the results due to the fact that they were held constant across the 6 trials.Recall Trials Both recall and recognition accuracy was analyzed for the pictorial stimuli. We first measured the correctness of the picture with disregard to its action. Mean accuracy with picture recall without actions was not significantly affected by whether the pictures were named (58.3%), named and written (50.4%), or named and sketched (68.7%), F(2, 40)= 1.52. Secondly, we measured the accuracy of the picture recall including the actions performed with the picture. Mean accuracy of picture recall including the action was significantly affected by whether the pictures were named (54%), named and written (42.9%), or named and sketched (65.5%), F(2, 40)= 2.163. Sophisticated motor programs (Engelkamp & Zimmer, 1984) which involved sketching, or the richer information generated were not responsible for higher accuracy of sketched and named rather than named pictures. Accuracy of written and named, was not higher than that of named despite the participation of multimodal processes in writing compared to just naming. Although sketching is not an automatic task whereas writing is, accuracy of the written and named pictures was the lowest. Influence of Repetitive Recall on the SPT Effect Data of picture recall revealed a slight increase in recall repetition interaction with each trial F(5, 100)= 2.24. Mean accuracy was the lowest in R1 (56.3%) where mean accuracy slightly increased during each trial except for R5 and R6 staying constant (61.1%). (Figure 1). The interaction of action x trials was not significant, F(10, 200)= .417Figure 1 Data of picture recall including the correct action involved revealed an increase after R1 and R2 (50%) to R3 (54.8%), then a slight decrease after R5 (57.1%) to R6 (56.4%). (Figure 2). The interaction of action by trial was not significant F(10, 200)= .86.Figure 2
Recognition Recognition for the pictorial stimuli was also tested for accuracy with and without the action presented with the pictures. Mean accuracy for picture recognition without the action was not significantly affected by whether the pictures were named (92.9%), named and written (95.2%), or named and sketched (97.6%), F(2, 40)= .488. (Figure 3).Figure 3 Mean correct action recognition differed across actions, on the other hand, and was significantly affected by whether the pictures were named (54.8%), named and written (40.5%), or named and sketched (83.3%), F(2, 26)= 9.750. The named and sketched action recognition was significantly higher than for named, and named and written. (Figure 4).Figure 4
The results of this experiment show that the effects of introducing different combinations of SPTs and VTs over repetitions to subjects are that pictures that were named and named/sketched had a higher mean of recognition. This contradicts what we originally hypothesized would occur. We expected to see the retrieval performance higher for the combinations of name/write and name/sketch actions. The combinations of actions did not seem to affect the recall over retrieval repetitions factor. Recall of the pictures with their respective actions was lower than just recall of the pictures themselves. The majority of the pictures were accompanied by the action ¡¥name¡¦. The free recalls for these were lower than for named/ sketched and named/written pictures. Action recognition was best for named/sketched pictures. However, picture recognition was unaffected by the action executed. The generation effect positively influences the encoding of the pictures. The generation effect contributes to the depth of the encoding. The depth of the encoding is directly correlated to the performance of repetitive retrieval. To explain the decrease of performance among named pictures, we must consider the nature of how the brain executes certain tasks from one region of the brain and other tasks from other regions. ¡§Tasks that utilize very simple, or automatic motor programs and provide little contextual information are in the extreme left (e.g. speech, reading), with opposite tasks in the extreme right (e.g. sketching, sculpting). This framework accounts for discrepancies under various conditions and the explanations provided by both Englekamp & Zimmer (1984) and Backman & Nilssen (1984).¡¨1 In free recall of the pictures with their actions, repetitions did not seem to influence the accuracy of the actions over repetitive retrievals. In recognition, the accuracy of identifying pictures with their corresponding actions was higher than the free recall trials. The richness of encoding affected both the free recall trials and the recognition tests comparatively. To conclude, our data suggests that the power of a combinatory set of actions that utilize the generation effect and deeper encoding processing provides for an increase in memory performance. Future research should explore the effects of having subjects use the richer encoding techniques of writing and sketching by performing these actions on half self-generated images and half presented stimulus. What would the recall performance be if the set of actions were the same for all images, but the order of pictures would alternate between self-generated images and presented stimulus? It is clear that the richer the encoding process, the more information can be retrieved.
ReferencesAdams, J., (1967). Human Memory. New York, McGraw-Hill Publisher.Burns, D.J., (1993). Item gains and losses during hypermnesic recall: Implications for the item-specific-relational information distinction. Journal of Experimental Psychology: Learning Memory, and Cognition, 19, 163-173. Cohen, R.L. (1983). The effect of encoding variables on the free recall of words and action events. Memory and Cognition, 11, 575-582. Craik, F. I., & Tulving, E. (1995). Depth of Processing and the Retention of Words in Episodic Memory. Journal of Experimental Psychology: General, 104, 268-294.Duffy, S.A., (1986). Role of Expectations in Sentence Integration. Journal of Experimental Psychology: Learning, Memory, & Cognition, 12, 208-219. Dunning, D. & Stern, L.B. (1992). Examining the Generality of Eyewitness Hypermnesia: A close look at time delay and question type. Applied Cognitive Psychology, 6, 643-657.Engelkamp, J. (1997). Memory for to-be-performed tasks versus memory for performed tasks. Journal of Experimental Psychology: Learning, Memory, & Cognition, 25, 117-124. Engelkamp, J., & Dehn, D. M. (2000). Item order and information in subject-performed tasks and experimenter-performed tasks. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 671-682.Engelkamp, J., Zimmer, D., Helstrup, T. (2000). Pop-out into memory: A Retrieval Mechanism That is Enhanced with the Recall of Subject-Performed Tasks. Journal of Experimental Psychology: Learning, Memory, & Cognition, 26, 658-670.Loftus, G. (1976). Human Memory: the processing of information. Hillsdale, N.J.: L. Erlbaum Associates Publisher. Talland, G. (1968). Disorders of Memory and Learning. Harmondsworth, Penguin Publisher. Matlin, M. (2002). Cognition (5th ed.). Belmont, CA: Wadsworth.Marmolejo, G., & Krizan, Z. (2001). Does Massed Retrieval Repetition of Simple Actions Increase Their Memorability? (Unknown Format for Citing a Convention)Olofsson, U. (1997). Win some, lose some: Hypermnesia for actions reflects increased item-specific processing. Memory & Cognition, 25(6), 797-800.
Citations 1Cited from ¡§Does Massed Retrieval Repetitions of Simple Actions Increase their Memorability?¡¨ Gloria Marmolejo & Zlatan Krizan. Winona State University ¡V Rochester.