AbstractThe aim of this context dependent memory test was to determine whether changing environmental contexts caused significant differences in free recall of a 20 word list. The participants ranged in age from 18-60 with gender being distributed 7:13, female:male. The study is similar to Abernathy (1940) and Fernandez and Glenbergs (1985) previous studies.
The independent variable with the context in which words were recalled, the dependant variable being the word being recalled. There was no significant results found with U = 48.5, below the critical value 27. The hypothesis that participants will recall words significantly better in the context in which they heard the words is false. IntroductionBackgroundContext dependant memory is within the cognitive psychology theory.
Context meaning an environment that influences a person’s memory. Tulving (1974) used cue-dependant forgetting when referring to context dependant forgetting.Atkinson and shiffrin’s (1969) multi storage model describes how three categories of memory storage work together; STM, LTM and sensory memory. Glanzer and Cunitz (1966) supports multi storage model.Memory is divided into short term and long term. LTM is visual, acoustic or semantic; and stored for later retrieval as LTM duration and capacity is far greater than STM. Jacob (1887) states STM capacity is between 5 and 9 items. Peterson and Peterson’s (1959) study influences the 30 second interference task in order to transfer words from STM to LTM.
Baddeley and Hitch (1974) suggest a multiple component STM, the working memory model, supported by Miller, Galanter and Pribram (1960). The phonological loop used for auditory recall includes the phonological store (inner ear), articulatory process (inner voice), and visuo-scratchpad (inner eye). Craik and Lockhart (1972) focused on depth of processing and encoding specificity; shallow processing (maintenance rehearsal) and deep processing (elaboration rehearsal). The questionnaire used highlights the different recall methods used. AimContext dependant memory refers to the idea that memory recall is enhanced within the environment that information was first learnt. Bower (1967), Tulving (1974)(1983), Anderson (1983) found advantages to contextual consistency within context dependant memory research.
To test recall vs recognition you would need to test a high number of words within a different study procedure.Abernathy (1940) found a 12% difference in recall within a classroom after a one week delay before recall. Saufley, Otaka and Bavaresco (1985) replicated this test with a sample of 9 participants, finding non significant results. Baddeley (1975) repeated the test using land and water as context environment, finding a 32% difference, however this is not comparable to the study being conducted. This study was replicated using a shorter inference time and different context, using a living room and dining room as opposed to classroom. Fernandez and Glenberg (1985) tested by recognition as opposed to recall, finding a 2% difference.HypothesisParticipants who learn and recall a word list in the same context will have significantly better results than those who recall in a different context from which they learnt the information.Due to past research stating the expected direction of correlation it is a one tailed hypothesis.
MethodDesignThe independent variable was the context in which information was recalled. The dependant variable was the number of words recalled. Participants undertook one set of experimental conditions, not both. Extraneous variables were controlled, such as outside noise, that may interrupt the test and make the results untrue.The word list included familiar words, avoiding alliteration or associations to ensure equal memorability.
A 30 second interference task was used to change the words from STM to LTM. A living room was used as the learning context and a dining room was used for the recall context. The allocation of participants to each context was opportunistic. The questionnaire filled out by participants highlights differences in age, gender, learning difficulties and language.EthicsEthical issues such as anonymity, right to withdraw and confidentiality were addressed before the test when participants signed a consent form (Appendix A). Consent was incomplete as participants needed to be naive of the test method in order to obtain genuine results. Debriefing explained this to the participants.
ParticipantsParticipants range from 18-60 years old, with a 7:13 distribution of women to men. The questionnaire outlined age, gender, learning difficulties, native language and recall method. Procedure20 participants sat in one context, a sitting room, and had a list of 20 words read out loud to them with a speed of approximately 2 seconds per word. Participants were then asked to list as many disney films as possible in 30 seconds as an interference task.10 participants were taken to a different context, a dining room, while the other 10 were asked to stand up and stretch to equate for the different context groups movement. Both contexts were then given 3 minutes to recall, in no order, as many of the words as possible. Throughout this time participants were asked not ot discuss between themselves.
After completing the recall participants were debriefed, Participants were asked to fill out a questionnaire, and any questions on the experiment were answered. Summary of ResultsMin WordsMax WordsRangeMeanPercentageSame Context (UA)716910.552.5%Different Context (UB)816810.351.5%When finished each participant was scored out of a possible twenty and scores were displayed for analysis (Appendix D). The mean for the some context was 10.
5, the mean for different context was 10.3. Ranges for same and different context were 9 and 8 respectively.
Same context minimum score was 7 and maximum was 16. Different context minimum score was 8 and maximum was 16. The Mann Whitney test was used to compare contexts.
With U = 48.5 and a critical value of 27 for one tailed experiments the results were found to be non-significant. Thus the hypothesis was rejected.The percentage difference between contexts was 1%, therefore non-significant. The same context had a 52.5% recall, different context had a 51.
5% recall. Because the difference is not significant it rejects the hypothesis that recall is higher when context is consistent. The 30 second interference task using in the study may not have been long enough to push the words from STM to LTM.Chart AnalysisFemale participants on average recalled 10.7142 words, mele participants recalled 10.2307. Dyslexic participants on average recalled 10.6666 words, non-dyslexics recalled 10.
3526.Native english speakers on average recalled 10.4117 words, non english speakers recalled 9.3333.There was no significant correlation between age and words recalled.
The pattern of words recalled supports Glanzer and cunitz (1966) serial position effect, which supports Atkinson and Shiffrin’s (1968)(1971) multi store model. DiscussionOutcome of FindingsThe hypothesis stated that recall would be significantly better when the context is kept consistent between learning and recall. Using a critical value of 27 and U = 48.5, The results found a non-significant 1% difference between context recall scores. Therefore the hypothesis was rejected. The percentage difference found is similar to results found by Fernandez and Glenberg (1985) of 2%. There was no apparent anomalies, with scores ranging 7-16 and 8-16 for same and different context respectively.
The questionnaire answered by participants highlighted difference such as learning difficulties, scores appear to be slightly for those with learning difficulties. Links to Past ResearchPast research tends to publish experiments with more significant results, which gives an unfair representation of tests such as context recall. Greenwald (1993) and Kruschke (2013) suggest this representation may cause researchers not to publish non-significant results as they are not seen as relevant to contextual memory research. Fernandez and Glenberg (1985) found non-significant results in a memory test. Smith and Vela (2001) while analysing previous research (1986) argued that environmental cues are encoded in memory, and using to provide cues suggested by Tulving (1974) to assist recall. Therefore changing context may be detrimental to recall, however this is challenged by the non-significant difference in the test. Abernathy (1940) found a significant 12% context difference, with a week between learning and recall, Possibly making the 30 second interference task unrealistic.
Otaka and Bravesco (1985) replicated the context dependant memory test using a small sample, encouraging anomalous results due to unrealistic generalisation of the sample group.Limitations and ModificationsMiles and Hardman (1998) states that different psychological states can affect context dependant memory, have a significant effect on clarity and method of recall, putting limitations on the test. If the participant is relaxed as a result of the context recall is expected to increase. Dzulkifli (2013) found that humour also has significant effects on recall, supported by Carlson (2011).The test design has limitations, by reading the word list as opposed to displaying it cause inhibit participants whose first language isn’t english, as well as limiting visual memory.Implications of Further ResearchThere are possible gender issues as a result of the disproportionate number of men. Herlitz, Nilson, Backman (2011) found that visuospatial memory is enhanced in men, while verbal capacity is enhanced in women. Further research into different recall methods such as repetition: Ebbinghaus (1913), Imagery: Paivio (1971), Loci: Bower (1972), Association: Pavlov (1927) and Narrative: Bower, Clark et al (1969), is critical in determining which method is affected by context.
Tulving (1985) and Anderson (1974) relate procedural memory to the narrative method of encoding STM to LTM. Acheson, Gresack and Risbrough (2014) while conducting PTSD and contextual fear learning research found links between encoding, retrieval, dysfunctional hippocampus and contextual representation in PTSD sufferers. Context elements are encoded to LTM during trauma and are more likely to trigger fear responses in the future. A longer interference time task could be used in future research as past studies with longer interference task have found different results to those with shorter interference tasks.ConclusionThe hypothesis, that learning and recalling a word list in the same context environment will give significantly better results then recalling in a different context environment was rejected.