Danah Henriksen
9.24.03


Long Term Memory Experiment



The following reported experiment was designed as a test of long-term memory, based on both immediate and delayed recall of a list of twenty words.  In this case, I decided to run the procedure on two different subjects, one of which was previously tested and described in my short-term memory experiment.  Subject 1 was a 56 year-old male, who in preceding short-term trials had demonstrated a relatively normal memory aptitude, in the sense that he had clearly demonstrated Miller’s (1956) “seven plus or minus two” rule of short-term memory.  I felt that it would be interesting to see whether he demonstrated some of the other principles or ideas described by Ashcraft (1989), with respect to long-term memory.  Subject 2 was a 53 year-old female, readily available to perform the memory trial immediately following Subject 1, who provided an interesting comparison in results.  Sets of data on experimental results for both individuals are provided below in Table 1 and Table 2.


On the immediate recall task, Subject 1 displayed an extremely impressive recall of 18 out of 20 words in the listing.  This score of 90% accuracy seemed exceptionally high, in particular as compared to the score of Subject 2, who recalled 13 out of 20 words for an accuracy rating of 65%.  Given that all other conditions for presenting the material was held constant, I presumed that some of the disparity between recall scores may be attributable to individual differences between the two subjects, in terms of innate memory ability and selective use of different mnemonic devices. 


One of the most significant findings that this study presented to me was its validation of the primacy effect.  It is clear from review of both individuals’ scores that words from the early portion of the list were recalled more quickly and easily than any others.  Despite the fact that there appears to be little similarity in the meaning of the words paper, seat, tire, and love, this initial set of words was uttered with speed and accuracy in both subject cases.  Ashcraft (1989) discusses the work of Rundus (1971), who concluded that rehearsal was the critical factor that the primacy effect depended upon.  In this experiment, which allowed for ten full seconds between words, the very first words repeated offered the greatest amount of time for rehearsal and were most easily committed to memory.  Additionally noted by Ashcraft (1989), was the fact that increasing the time for study between items also increased recall of earlier items, but not the end of the list recall.  Both of my subjects validated this point clearly and concisely, demonstrating serial accuracy with the initial words given, and struggling a bit both with recall and order in later words.  


Generally speaking, it is safe to say that rehearsal played a significant role in the long-term memory effect of this experiment.  The ten seconds given between words provided subjects with additional time to rehearse the information thus committing it to long-term memory with the likelihood of improved performance in later memory trials.  Ashcraft’s (1989) discussion of
Atkinson & Shiffrin (1968) expressed this occurrence as such that rehearsal allows for information to be held longer in short-term memory, thus increasing it’s chance for being assigned to long-term memory and permanent storage.  As such the short-term experiment from last week did not provide ample time (only one second) for information to be rehearsed within short-term memory.  Thus, performance was never raised beyond the “seven plus or minus two” rule (Miller, 1956), in which little more than approximately seven pieces of information can be temporarily encoded and retrieved from short term memory. 


The delayed recall exhibited by my subjects was not a reflection of the forgetting curve described by Ebbinghaus (1885).  This phenomenon predicts that the majority of “forgetting” of information in long-term memory occurs early, with a significant drop in recall within the first hour.  However both subjects, though they did struggle a bit more after an hour of time, were able to retrieve the majority of the information they had originally committed into long term memory.  This fact is perhaps more reflective of several factors noted by Ashcraft (1989) in the current model, which refute some of Ebbinghaus’ principles.  Namely, these refuting factors are, that subjects invent meaning as created in their memory, and that subjects are active participants who apply mental resources and strategies to virtually every learning situation (Ashcraft, 1989).  In both cases, my subjects were instinctively applying active effort and mental resources to commit the experimental list to memory.  Through their own processes of creating meaning and mentally rehearsing words and ideas, they were able to retain a majority of their original list recall, even after an hour after the immediate recall test. 


As an aside, I mentioned earlier that individual differences between people seem to be influential towards performance on memory task, and made for an interesting (almost entertaining actually) note in this study.  The subjects each approached the task quite differently, which seemed to be reflected in recall success.  Subject 1 applied himself with studied and intense concentration, appearing silent and remote during the initial reading of the list.  Subject 2 however, displayed much greater anxiety through the task.  This anxiety was evidenced by several distraught comments over her declared inability to remember anything, as well as an exclamation, halfway through the list, that she had “forgotten it all”.  Given this, I was surprised and impressed that she recalled even 13 out of 20, and wondered if a less anxious approach might have improved her score further.  Per my own observation and some discussion after the final recall, it seemed that each subject also opted for quite different mnemonic devices, based upon their instinctive understanding of how they learn, and devices they had always employed.  Subject 1 employed more of a narrative approach to the words, creating sentences or phrases in her head to string the words together.  For example she stated (with complete seriousness) that phrases like “Chairman Mao walked into the woods carrying his green gift” were helpful to her in recalling words.  Therefore for this individual, a semantic approach to the words themselves seemed to be the most natural method, and may be reflective of a cognitive style in which she focuses on words and narrative meaning, as opposed to concentrating on visual symbols.  Subject 2 however, found the use of visual symbols to be most salient in recall tasks.  When asked about his use of memory devices, he noted a strategy of “visually building symbols on top of each other in his mind”.  (i.e. placing a tire and a seat on a piece of paper, etc.)  In discussing this he also noted that it perhaps caused him to struggle more with words such as “analysis” for which it is difficult to call up an image to mind.  In particular Subject 2’s use of imagery corresponded well to an interesting point made by Ashcraft (1989).  The dual coding hypothesis proposed by Paivio (1971) “words that denote concrete objects, as opposed to abstract words, can be encoded into memory twice”, on both the basis of verbal and imaginal attributes.  This was helpful in explaining Subject 2’s ease of recall for tangible objects, given his cognitive approach to the task. 


In general, these observations of general principles of memory as well as individual differences in approach to the learning/memory task demonstrate the impact that multiple factors have upon long-term memory.




Table 1 - Subject 1 Results

Actual List Data

paper, seat, tire, love, beach, analysis, conjunction, brush, chairman, accurate, woods, green, hunger, gift, keyboard, number, bottle, jogging, wheel, system.

Immediate List Recall

paper, seat, tire, love, conjunction, chairman, accurate, woods, green, hunger, keyboard, number, analysis, bottle, wheel, number, system, brush

Immediate Recall Score

Non-serial recall of 18/20 with 90% accuracy

Serial recall of 14/20 with 70% accuracy

Delayed List Recall

paper, seat, tire, love, beach, analysis, chairman, brush, accurate, wood, green, hungry, bottle, wheel

Delayed Recall Score

Non-serial recall of 13/20 with 65% accuracy

Serial recall of 11/20 with 55% accuracy



Table 2 - Subject 2 Results

Actual List Data

paper, seat, tire, love, beach, analysis, conjunction, brush, chairman, accurate, woods, green, hunger, gift, keyboard, number, bottle, jogging, wheel, system.

Immediate List Recall

paper, seat, tire, love, beach, conjunction, analysis, chairman, woods, green, keyboard, bottle, system.

Immediate Recall Score

Non-serial recall of 13/20 with 65% accuracy

Serial recall of 11/20 with 55% accuracy

Delayed List Recall

paper, desk, love, beach, wheel, jogging, bottle, keyboard, system, woods, green, chairman, gift

Delayed Recall Score

Non-serial recall of 12/20 with 60% accuracy

Serial recall of 6/20 with 30% accuracy



Ashcraft, M. (1989).
Human Memory and Cognition (pp. 187-245). Scott, Foresman and Company: Glenview, IL.