Remembering False Words - Essay Example

1Running Head: Imperfect Memory: Remembering Imperfect Memory: Remembering False Words William Council In Roediger and McDermott's 1995 experiment, students remembered a false word from a list of 15 semantically related words. After memorizing lists of 15 words, the participants in my experiment were shown three words - 1 word from the list, 1 word visually similar to the list and 1 word unrelated to the list. Using a scale from 1 to 4, participants measured how sure they were that each word was listed earlier. Participants "remembered" visually and semantically similar words more often than unrelated words. Introduction Cramming for test is a recipe for failure. There are times when the brain's associative capacities are a hinderance for accurate recall. Roediger & McDermott's (1995) had Rice University undergraduate students review six lists. The lists were drawn from listed in Deese's 1959 article. After the students reviewed the list, they were presented with another list of words. They were asked to identify the words that were on the first word list. Of the 42 words, 12 were from the original list and 30 were word lures. The 30 word lures fell into three categories: 6 words that inspired the original word lists, 12 words unrelated to the earlier list and 12 words slightly related to the list. The Roediger & McDermott experiment focused on semantically similar words. Sematically similar words are words that share an association. For example, bed is semantically similar to sleep and night. Roediger & McDermott's results noted that words that appeared visually different than list words were not falsely remembered. Subjects distinguished these false words easier than false words that resembled list words. This experiment also used three types of word lures. Instead of using a group of slightly related words, visually similar words were added. Visually similar words were tested to see if subjects remembered the image of the word as well as its meaning. Visually similar words look alike. For example, blade is visually similar to blank. This experiment focuses on three premises: 1. Semantically similar words will be falsely remembered more often than unrelated words, 2. Visually similar words will be falsely remembered more often than visually unrelated words and 3. Semantically similar words will be remembered more often than visually similar words. Method Participants: 101 ARU first year psychology students. Materials: The six semantically similar word lists were based on the lists used in the Roediger & McDermott study. A computer program was designed to create the visually similar word lists. From a list of 12000 words, the program chose ones that fit the following criteria: a). At least 5 characters long, b). Words do not have common English endings (ing, ed, s) and c). Each word must have at least 15 other words that differ from it by no more than 2 letters. Design: Both the visually and semantically similar lists were refined. The semantically similar list was refined to reflect different types of word associations. Words in the same category (chair and bench), common word combinations (needle and haystack), opposites (mountain and valley), adjective to noun (rough ground) and noun to adjective (needle knitting). From the computer generated visual list, six word groups were hand selected. These groups were required to meet additional criteria. First, the finalists had a common and concrete target word. Secondly, the final word groups did not include words that all sound alike or had the same first letter. Six word lists were chosen for the visual test set. Procedure: Participants are asked to memorize a list of 15 words. After each list, the participants are presented with three words. One word is from the earlier list. The second word is one unrelated to the list. The third word is either visually or semantically similar to the word list. Participants indicate how confidant they are that the word is from the earlier list. They use a scale of 1 to 4: 1 meaning the word was definitely new and 4 meaning the word was definitely on the earlier list. Results The mean for the semantic target word was 2.3949 with a standard deviation of .76108. The actual semantic word with a mean of 3.2354 and a standard deviation of .56394. The unrelated semantic word's mean was 1.1574 and a .31297 standard deviation. The mean for the visual target word was 2.1965 with a .69638 standard deviation. The actual visual word received a 3.1451 mean and a .53127 standard deviation. The visually unrelated word had a mean of 1.1838 and a .40372 standard deviation. The results supported the three hypothesises. Participants incorrectly remembered the semantic target word more often than an unrelated word. (2.3949 mean to 1.1574) Visually target word experienced a similar victory over visually unrelated word. (2.1965 mean vs 1.1838) Semantically target words were incorrectly remembered more often than visual target words. (2.3949 vs. 2.1985) Discussion This study shows that subjects will incorrectly remember words as being on semantically and visually related word lists. The findings support all three hypothesis including the one stating that semantically similar words create a higher rate of false memories than visually similar words. This suggests that some word associations are stronger than visual similar words. An interesting follow up experiment would to have the lists of words read to the subjects while showing them on a screen. A recent experiment examined the relationship recollecting studied items and incorrectly remembered items. (Geraci and McCabe, 2006) Adding the visually similar words to a new test that includes facets of Geraci's experiment would continue to gather data on the role visually similar words play in accurate short term memory recall. Repetition, most oftenly used when students either cram for a test or attempt to learn a second langauge, could also be examines from the visually similar word list angle. (McDermott, 2006) Greater understanding of the short term memory illusionary recall phenomena would assist educators, students and aging baby boomers. Many people make lists and read them aloud with hopes of remembering the material whether it's groceries, phone numbers or the elemental table. This experiment and others expands our understanding of short term memory and brings us closer to improving it. Other testing could examine the relationship between short and long term memory. Short term memory errors are more acceptable in today's society. Numerous gadget store information that used to be entrusted our short term memory. Recent experiments found that semantic word associations affect short term memory but not long term memory. (Davelaar et al., 2006) The focus for the new testing would be examining the visually similar results in the long term section of Davelaar's work. This experiment expanded on Roediger and McDermott earlier work. (1995) The results from this experiment and other recent works lay a foundation for future short term memory experiments. A future experiment that include auditory and repetitive recall elements is the next logical step. References Davelaar, E.J. & Goshen-Gottstein Y. & Haarmann, H. J. & Usher, M. (2006). Semantic similarity dissociates short-from long-term recency effects: Testing a neurocomputational model of list memory. Memory & Cognition, 34(2), 323-334. Retrieved May 8, 2008, from Alumni - ABI/INFORM Complete database. (Document ID: 1049816331). Geraci, L. McCabe, D. P. (2006). Examining the basis for illusory recollection: The role of remember/know instructions. Psychonomic Bulletin & Review, 13(3), 466-73. Retrieved May 8, 2008, from Alumni - Research Library database. (Document ID: 1141331851). McDermott, K. B.(2006). Paradoxical effects of testing: Repeated retrieval attempts enhance the likelihood of later accurate and false recall. Memory & Cognition, 34(2), 261-267. Retrieved May 8, 2008, from Alumni - ABI/INFORM Complete database. (Document ID: 1049816271). McDermott, K. B. & Roediger III, H. L. (1995) Creating false memories: remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21(4), 803-814. Read More