Eastern Oregon University
General Education Assessment
2008 Spring Pilot Project:
Critical Thinking Outcome
English 335: Cognitive Psychology
Student Samples
Level 3: Proficient
Running head: REMEMBER AND KNOW PARADIGM
Remember and Know Paradigm in the Recognition Memory Test
Eastern Oregon University
Abstract
In cognitive psychology, remembering and knowing are distinct. Remembering involves contextual recollection, while knowing involves feeling. To investigate this topic, I did an experiment in which 10 university students observed 80 words and either wrote down a synonym or a word that rhymed with the original, depending on the instructions that appeared on-screen. Next, they were shown the first 80 words and another 80 words not presented before. The task was to judge whether a word was presented in the first group of 80 or not. If yes, they were asked to judge whether they consciously remembered or just knew. The results showed that synonym words helped people consciously remember, but the effect couldn’t be seen in know judgments.
Remember and Know Paradigm in the Recognition Memory Test
How clearly do we remember experienced events from our past? Our memories can be inaccurate. When we remember some events, it is not uncommon to not be sure whether we are consciously aware of them or we just unconsciously know them. What is a difference between conscious awareness and unconscious awareness? Many researchers have investigated the question. The most famous researcher who investigated the area of memory recognition was Tulving, who introduced the Remember/Know paradigm (Tulving, 1985, as cited in Brandt, Gardiner, & Macrae, 2006). The basic Remember/Know recognition memory task required participants to answer whether the item shown to them was old or new, and if old they were further asked if they were consciously aware of the item or just knew it. If they were consciously aware of the content of the item, it meant they remembered it, but if they were not, it means they know it (Brandt, Gardiner, & Macrae, 2006). The purpose of the study was to discover how we recollect events. Based on the findings, Tulving suggested that remembering related to our conscious awareness, and knowing related to our unconscious awareness (Brandt, Gardiner, & Macrae, 2006). Moreover, remembering has something to do with the episodic memory system, the one involved with our personal experiences, and knowing has something to do with the semantic memory system, the one concerning with abstract facts and knowledge (Goldstein, 2008). Based on his idea, when remembering personal events, we are likely to use mental time travel, a way of remembering events by going to the past in spirit, as it were (Goldstein, 2008). With this idea, Gardiner (1988, as cited in Brandt, Gardiner, & Macrae, 2006) suggested that there was some relationship between remember/know and subjective experiences and distinctive information.
To figure out how distinctive events affect remembering and knowing in a recognition memory test, Brant, Gardiner, and Macrae (2006) did an experiment using distinctive and typical forenames. The participants were 34 undergraduates and they were presented with 80 target forenames (40 distinctive and 40 typical forenames). They were first presented with the 80 target forenames and then the order was counterbalanced. After the first phase, they were presented with the 80 target forenames and an additional 80 forenames on the computer. The task was to make old or new judgments for each forename. If they thought the forename was presented before, they said it was old. If not, they said it was new. When they answered that the forename was old, next they had to identify their recognition of the forename as either ‘remember,’ ‘know,’ or ‘guess.’ If they were consciously aware of the forename and the content, they made a remember judgment, while if they couldn’t consciously recollect the forename and just knew that the one was presented before, they made a know judgment. If they had no idea whether they remembered or knew, they made a guess judgment (Brandt, Gardiner, & Macrae, 2006). According to the results of the experiment, distinctive forenames were more memorable for the participants compared to typical forenames. The conclusion was that distinctive events or information are more likely to help people consciously remember things (Rajaram, 1996, as cited in Brandt, Gardiner, & Macrae, 2006).
To review and investigate this topic, I did an experiment similar to the recognition memory test. The experiment involved 10 university students (6 women and 4 men). I assigned them to judge words as ‘remember’ or ‘know’ like the previous experiment, but this task required them to write a synonym or rhyme word that relates to the target words presented on the computer. I encouraged them to use the written words as a memory cue. This experiment related to levels-of-processing theory, a theory that suggests that how we encode information has an effect on the accuracy of our memory (Goldstein, 2008). Generally, there are two types of processing, deep processing, involving meaning, and shallow processing, involving no attention to a word’s meaning. When we encode information using deep processing, our memory is more accurate than by using shallow processing, according to several studies (Goldstein, 2008). In terms of this experiment, writing down a synonym involves deep processing because the word has the same meaning while writing down a rhyme word refers to shallow processing because it requires no attention to meaning. The main purpose of the experiment was to see how the type of processing, whether deep or shallow, affected their recollection of the target words. I predicted that synonym words are more likely to trigger more accurate remember and know judgments and rhyming words to trigger more accurate know judgments.
Method
Participants
The participants were 10 university students (6 women and 4 men) from Eastern Oregon University who took a cognitive psychology class. They volunteered for the experiment as a class exercise.
Materials
The computer software, called CogLab, was used to do the experiment, and the participants prepared a pen and two pieces of paper by themselves before the experiment (Vanhorn, 2008). There were two phases. In phase 1, the participants were presented with 80 words (such as chair, book, love, want, etc.) on the computer screen. In phase 2, they were presented with the first 80 words and an additional 80 words that were not presented in the first phase (Vanhorn, 2008).
Procedure
In phase 1, the participants were shown a word with an instruction that said either “Synonym for” or “Rhymes with” (Vanhorn, 2008). If the instruction said, “Synonym for”, they needed to write down another word on paper that had the same meaning as the word shown on the computer. For example, if the presented word was “power”, then they wrote down a word like “force.” If the instruction said, “Rhymes with”, then they needed to write down a word with which the target word shared a similar sound. For example, if the target word was “book”, they would write down a word like “hook.” Each target word was shown on the screen for only 5 seconds, so they needed to write as quickly as possible. In the second phase, they were shown the 80 words shown in phase 1 and the 80 new words that weren’t shown in phase 1. When one of the total 160 words was presented, the instruction said, “Was this word in phase 1?” and if they decided that it was, they needed to judge whether they remembered or knew the word. To “remember” means that they were consciously aware of the word and the related event that happened in phase 1, so that they remembered it with confidence, but if they couldn’t remember accurately or they just felt the word had been presented in phase 1, they chose “know.” When they remembered the word, they needed to push the r key on the keyboard, and when they knew it, they needed to push the k key. Finally, if they thought that the word was new and not presented in phase 1, they pushed the n key. After the two phases, the accuracy for each of the three categories (“remember,” “know,” or “not on the list” choice) in the two conditions (Synonym and Rhyme) was measured.
Results
The average accuracy for “remember” judgments was 0.61 in the synonym condition (SD = 0.28) and 0.505 in the rhyme condition (SD = 0.23). The average accuracy for “know” judgments was 0.20 in the synonym condition (SD = 0.25) and 0.26 in the rhyme condition (SD = 0.2) (see Figure 1). Finally, the average accuracy for “not on the list” was 0.19 in the synonym condition (SD = 0.14) and 0.23 in the rhyme condition (SD = 0.14). There was no significant difference between the average accuracy of “remember” judgments in the synonym and rhyme conditions, t (9) = 1.85, p = 0.098, as tested by a paired two sample t-test, but the mean accuracy for the remember judgment in the synonym condition was comparatively higher. Also, there was no significant difference in mean accuracy of “know” judgments between the two conditions, t (9) = 1.48, p = 0.17, as tested by a paired two sample t-test, but the mean accuracy for the know judgments in the rhyme condition were comparatively higher.
Figure 1. Means of the proportion of correct recognition of “remember” and “know” judgments in the synonym and rhyme condition. Each line indicates the relationship between the two conditions.
Discussion
As I predicted based on the levels-of-processing theory, the mean accuracy for “remember” judgments in the synonym condition was higher than that in the rhyme condition because when they encoded the information about words, they wrote down the synonym word that had the same meaning as the target word (see Figure 1). The synonym word became a trigger for the participants to consciously remember the target words. Therefore, LOP (levels-of-processing) effect was shown in “remember” judgments.
On the other hand, the mean accuracy for “know” judgments showed results opposite to the “remember” judgments. The mean accuracy in the synonym condition was lower than that in the rhyme condition, so LOP effect wasn’t shown in this case (see Figure 1).
The results were as I predicted, but why did the results show that the mean accuracy for the remember judgments and the know judgments in the two conditions was different? One possibility is that the know judgments might depend on another memory system. As I said before, one possible explanation about this question is that remember judgments relate to the episodic memory system, but the know judgments relate to the semantic memory system (Vanhorn, 2008). Probably, the know judgments were involved with procedural memory system, a type of implicit memory (Goldstein, 2008). The procedural memory means that we unconsciously know how to do things such as swimming, throwing a ball, and so on. This memory often refers to some skills that remain after we first learn those activities (Goldstein, 2008). Therefore, the reason why the mean accuracy for the know judgments in the rhyme condition was higher may be that we encoded the target words by using the procedural memory system rather than the episodic memory system. According to Gardiner and Java (1993), remembering is to recollection of content as knowing is to feeling or familiarity (as cited in Brandt, Gardiner, & Macrae, 2006). Thus, the mean accuracy for the remember judgments in the synonym condition was higher compared with the rhyme condition, and the mean accuracy for the know judgments in the synonym condition was lower compared with the synonym condition.
The collected data on the experiment seem reasonable, but there are still limitations to the experiment. First, each participant did the task individually, not in any laboratory. Nobody could observe that they correctly did the task and followed the directions. For example, they were asked to write down a synonym or rhyme word that related to the target word. What if they wrote down a word that was exactly the same as the target word without following the direction? The recognition memory test was another part with possible problems. What if participants just copied the target words, so that they could very easily answer whether the words presented in phase 2 were presented in phase 1? The results would be untruthful. Next, the judgments about whether they remembered or knew the words depended on the participants’ self-judgment. Their self-judgment might be unclear. What is an accurate borderline between the remember judgment and the know judgment? It depends on their self-judgment. So, the borderline should have been identified by other measures such as forced-choice question so the experiment would have been more objective.
Despite the limitations, the results showed that the mechanism of how we encode information has something to do with our memory recollection. If we use deep processing when encoding (the synonym condition), it helps us consciously remember the content of the information whereas if we use shallow processing (the rhyme condition), it often doesn’t help us remember the content of the information but help us know the information thanks to an idea of familiarity. The experiment is useful in helping us to know the mechanism of the levels-of-processing. Based on the experiment, I would like to research how emotional informational words affect remember and know judgments as future research involving the recognition memory test.
References
Brandt, K. R., Gardiner, J. M., & Macrae, C. N. (2006). The distinctiveness effect in forenames: The role of subjective experiences and recognition memory. British Journal of Psychology, 97, 269-280.
Goldstein, E.B. (2008). Cognitive psychology: Connecting mind, research, and every day experience (2nded.). Belmont, CA: Thomson Wadsworth.
Vanhorn, D. (2008). CogLab online manual (2nd ed.). [Computer software]. Belmont, CA: Thomson Higher Education.
Level 2: Adequate
Running head: TO REMEMBER OR KNOW
To Remember or Know: A Psychology Class Exercise
Eastern Oregon University
Abstract
Do people remember specific things easier when they have the opportunity to associate it with something else like it? A study was conducted with 10 college students who were asked to preview a list of words and then match each word with another word that either rhymed with it or was a synonym for it (phase 1). Afterwards the participants were shown the same list of words plus new words and were asked to decipher if each word was shown on the first list or not. They were asked to respond to each word as if they “remembered” or just simply “knew” it was on the first list. The results showed that the participants were more efficient on correctly identifying a word, which previously had a synonym for it, was truly on the first list when they responded that they “remembered” it. In conclusion, when people are trying to remember something, they remember more efficiently when they associate it with something else similar to it, in terms of meaning.
To Remember or Know: A Psychology Class Exercise
There are many ways scientists believe how daily information is turned into memories. A level of processing (LOP) is the idea that there are different depths of processing in which information can be encoded. One type of processing is deep processing. Deep processing involves attention to meaning and takes in-depth rehearsal. A second type of processing is shallow processing. Shallow processing involves repetition with small amounts of attention. Another type of processing is fluency processing, which is known as “a sign of prior experience and therefore serves as a cue for recognition” (Kurilla & Westerman, 2008, p.82). Many people’s everyday decisions are influenced by their interpretation of factors that could possibly give rise to fluent stimulus processing (prior stimulus exposure) (Kurilla & Westerman, 2008).
Kurilla and Westerman (2008) conducted an experiment to test the notion of the fluency affect. They conducted a similar experiment to a remember/know paradigm. In a typical remember/know paradigm, “participants are required to introspect about the nature of their memorial experience during a recognition memory test and to categorize each positive recognition response as one of either remembering or knowing” (Tulving, 1985, as cited in Kurilla & Westerman, 2008, p.83).
In order to “remember” a test item, the participant must have been able to have a though/association that directly linked to that test item from the prior study phase. And in order to “know” a test item, the participant must have been confident that the item was from the prior study phase, but he/she couldn’t exactly point out specifics about it (Kurilla & Westerman, 2008). The response, to “remember” a test item, requires levels of processing, unlike the response, to “know” a test item.
For Kurilla and Westerman’s experiment, they predicted that “manipulations of perceptual fluency primarily influence claims of knowing rather than claims of remembering” (Kurilla & Westerman, 2008, p.84). They tested 70 college students who participated in being presented with test items. The test items consisted of 60 words, which half of those words were previously studied and the other half of the words were new. The participants reported “yes”, “remember”, or “know” responses to the test items, which they thought, had been previously studied.
Their results showed that the previously studied words were primarily observed in “know” responses, with little influence on the “remember” responses (Kurilla & Westerman, 2008).
In the current study, participants were challenged with a similar remember/know task. The task was designed for participants to respond that they “remembered” or “knew” a word that was shown in the second list, was also shown in the first list. The first test list of words also had previous matched synonyms and words that rhymed with them. I predicted that the participants would respond with more “remember” synonym responses than with the rhyme response or any “know” responses, just like Kurilla and Westerman had hypothesized for their experiment.
Method
Participants
Ten college students (4 males and 6 females) of the same cognitive psychology class volunteered to participate in a remember/know class exercise. Before the participants began the exercise, each one of them read and signed an informed consent form. All participants were treated equally and ethically.
Materials
There were two phases included in this experiment. In Phase 1 there were 80 words and in Phase 2 there were 160 words. The types of words varied. This experiment was conducted on a computer, online. Each participant needed to have a blank sheet of paper and a writing utensil.
Procedure
The participants were asked to log on to the website (coglab.wadsworth.com) with the remember/know task. They were instructed to have a blank sheet of paper and writing utensil available for Phase 1. To start each phase and/or trial, they were to press the ‘space bar’ key. For Phase 1, there were 80 words. Each trial a word and an instruction for that word appeared on the screen. Each participant was to follow the instruction given. There were only two types of instructions for Phase 1. One instruction simply asked the participant to write down a synonym for the particular word and the other instruction asked the participant to write down a word that rhymed with the particular word. The synonym instruction, which takes deep processing, was designed to have the participant write down a word with similar meaning and the rhyme instruction, which takes shallow processing, was designed to have the participant write down a word with similar sound. Each word and instruction was only viewable for five seconds. After the participants finished Phase 1, they went directly into Phase 2.
Phase 2 consisted of 160 words and one question. For each trial, a word and the question, “Was this word in Phase 1?” would appear on the screen. Each participant was to press the ‘r’ key if he/she “remembered” the word being in Phase 1, press the ‘k’ key if he/she “knew” the word was in Phase 1, or press the ‘n’ key if the word was not included in Phase 1.
In Phase 2, half of the 160 words were actually in Phase 1 and the other half of the words were not.
Results
As predicted, the participants correctly identified a word shown in Phase 2, also being shown in Phase 1, more efficiently by responding that they “remembered” it compared to the participants’ response of “knowing” it. The average proportion correct across participants for “remembering” a word with the instructions to write down a synonym for it was .61 and the instructions to write down a word that rhymed with it was .50, as shown in Figure 1. The average proportion correct across participants for “knowing” a word with the instructions to write down a synonym for it was .2 and the instructions to write down a word that rhymed with it was .26, as shown in Figure 1.
A paired t-test between “remember” word responses (synonym for and rhymes with) indicated that there really wasn’t a significant difference between words with synonyms and words with rhyming words, t (9) = 1.85, p <.098. A paired t-test between “know” word responses (synonym for and rhymes with) indicated that, again, there really wasn’t a significant difference between words with synonyms and words with rhyming words, t (9) = 1.48, p <.17.
Figure 1. Mean proportion of the “remember” and “know” response of correctly recognized words shown in Phase 2 that was also shown in Phase 1.
Discussion
The results supported the prediction that the participants would respond more correctly with the “remember” synonym responses compared to all the other options/responses. This conclusion, leads to the idea that in order to remember something, it’s best to associate what ever it is with something similar in meaning.
Kurilla and Westerman had similar predictions and results even though the experiments and variables differed. The idea of levels of processing holds true when compared to both experiments’ results. In order to “remember” an item a person must connect it with something of valuable meaning, because it takes the level of deep processing, which results in better performance of actual remembrance.
References
Kurilla, B.P., & Westerman, D.L. (2008). Processing fluency affects subjective claims of recollection. Memory and Cognition, 36, 82-82.
Level 1: Developing
Running head: CONSCIOUS/UNCONCIOUS RECOGNITION MEMORY
Conscious and Unconscious Relationship in Recognition Memory
Eastern Oregon University
Abstract
This experiment examined the conscious and unconscious relationship in recognition memory in an experiment where a group of 10 college students (six female and four male) were shown a list of 80 words in phase I and asked to generate either a rhyme or a synonym for each one. In phase II of the experiment the students were shown 160 trial words and asked to note whether the word was remembered, recognized, or not remembered. Although remembered and not remembered were absolute reactions to the words, recognized words allowed the students to respond based on an unspecified memory as it related to the phase I words. Examination of the test results shows that recognition by remembering was much higher than that of knowing absolutely they had seen those words. Unconscious memory recognition shows a higher correlation towards recalling words than conscious recognition.
Conscious and Unconscious Relationship in Recognition Memory
Unconscious memory recognition can lead to false memories which lead to erroneous and unreliable accounts of past events. Many people are affected by such memories which may or may not result in being accused of criminal activities and all together losing their families or loved ones. In our research study, general word recognition was much higher than actual word recognition. Words like other memories are often correlated to past events or learned through general experiences and then recalled on a broaden spectrum when related to a factual presentation. If the theory that unconscious word recognition is more common, it follow that this pattern holds true for other memories as well.
In view of the data collected in the experiment, I predict that, unconscious memory recognition plays a major role in false memory.
Method
Participants
Ten members of a cognitive psychology class (4 men, 6 women) participated voluntarily as part of a class exercise. All participants were treated equally in this experiment. It should be noted that two participants (1 male, 1 female) are ESL (English as a Second Language) students.
Apparatus
This experiment was conducted as an online CogLab. Each participant was required to do the experiment on a computer or laptop that had the Java application. The participant was allowed to take the test at their leisure and in the environment of their choosing.
Materials
The participants were required to have a piece of paper a pen or pencil to write down the synonym or rhyme requested by the test.
Procedure
Each participant was required to sign onto their CogLab account and select the Remember/Know lab. The participant was required to read the directions of the experiment which stated that they would be shown a list of 80 words and would be asked to write down a synonym or rhyme for the each word displayed. Once the participants were ready they were instructed to hit the space bar to start the experiment. Words appeared individually with instructions to write either a rhyme or synonym, staying on screen for a few seconds, and then progressing to the next one. Phase I continued until the entire list was completed.
Phase II began with instruction to the participant to enter the letter r if they remembered the word, k if they knew the word or, n if the word was not in the list during phase I. The judgment remember means that they were consciously aware of the learning episode. The judgment know means that you unconsciously aware of the learning episode but you knew the word was on the list. There were 160 trials shown in Phase II consisting of the Phase I list as well as words that were considered “Lures” (ie: words that were meant to trick the participants). Phase II proceeded with no time limit, subsequent words did not appear on the screen until a judgment was entered.
Results
Primary analysis focuses on the mean of the test group in proportion to correct recognition which is indicated in Figure 1. For the remember synonym test results were 61% correct and for the remember rhyme they got 50% correct. For the know synonym test results were 20% correct and for the know rhyme 26% correct. The two-sample t test indicated as followed t(9)=1.85, p < .098. Although the t test doesn’t show a significant difference, the results are going in the direction as predicted.
Figure 1.
Discussion
The results of this experiment show that in the case of word recognition, conscious word recognition was significant for this test. However, it should be noted that Phase I of the test was followed immediately by Phase II and may have shown a significant increase due to short term or even sensory memory recognition. Unconscious word recognition showed a relatively low proportion of correct recall. These results do show that when presented with an incorrect answer the mind can often have an incorrect response.
These results show that even on a simple task like word recognition, previously related memories can create false memories or the perceived thought that they had just seen or read a word in some context. Ira Hyman, Jr. conducted a similar experiment that related more towards real life experiences by creating false memories of past life events. He asked the participants parents to make up false memories such as having a clown at a birthday party or spilling a bowl of punch at a wedding reception. The results were shocking, 20 percent of the participants recalled and even described the events in detail that had been implanted into their false memory by their parents. Another correlating factor is eye witness testimonies used in criminal cases. As eye witness testimonies being one of the most compelling factors to a jury, only 29 percent are considered confident and accurate. As derived from our test results the percentage from both examples correlate with the test results on unconscious word recognition. This is significant because unconscious word recognition could be confused as a false memory which directly relates to the false life memories in the life events. Also it connects with the eye witness example because the eye witness’s testimony was not confident in the choices they made as well as the remember/know section of Phase II in the word recognition experiment.
False memories are generated when unconscious thoughts are used to replace actual events. The subject then has what is believed to be accurate recollection of events, people, places, or conversations that are incorrect or completely false. As in the conscious and unconscious word recognition experiment, subjects may believe that a word was presented to them, when it was not, thus creating a false memory in word recognition.
In the Functional aspects of recollective experience conducted by John M. Gardiner, recognition in the absence of conscious recollection, although less likely, was generally reliable and uninfluenced by encoding conditions. The results are consistent with dual-process theories of recognition and our word recognition experiment.
This study is an important step in understanding false memory. Analysis of the results of the conscious and unconscious word recognition experiment support, at least in part, the factors involved with false memory. Unconscious memory process may in fact pay a major role in false memory because if we could easily create a false memory involving a word than it is safe to conclude we could create false memories about people, places, events, and ext. Creating these false memories could be detrimental to the individual or people in their lives.
References
Gardiner, John M. Functional Aspects of Recollective Experience. Memory & Cognition (1988): 309-313. EBSCOhost. 20 May 2008.
Goldstein, Bruce E. Cognitive Psychology: Mind, research, and everyday experiences. 2nd ed. Belmont, CA: Thomson Wadsworth, 2008. 264-276.
Vanhorn, Daniel. CogLab Online Manual. Cognitive Psychology: Mind, reseach, and everyday experiences. Belmont, CA: Thomas Wadsorth, 2008. 87-89.