1 Abstract The experiment conducted regarding Memory Processes tested individuals on their ability to store and retrieve words. The levels on which words were stored were structural, the lowest level, phonetic, the next highest level, and semantic, the highest level of processing. The experiment is based on the recall and reorganization of the words from group they show during the experiment. The experiment conducted supported hypotheses regarding a subject's performance on retrieving words at different levels. The independent variables were the encoding levels, and they manipulated the dependent variables by affecting the time in which a word could be received due to its placement on a processing level. This experiment was an extension of Craik and Lockhart Depth of Processing Model, that explored the effects of "deeper" processing by an individual, and the likely-hood that those words processed deeper were retrieved better.
2 Introduction In 1970, Craik and Lockhart proposed that there are different levels of processing a person uses while encoding information. Thus, they devised a model to represent these levels of processing called the Depth of Perception Model. In this lab, the levels of processing were based on a person's ability to recall certain words according to the category in which they were presented to the subject. The three levels of these encoding categories were orthographic (structural) processing, or identified physical characteristics; phonetic processing, or the sound a word makes and that auditory relationship to other words; and semantic processing, or the representational meaning a word has when used in context with other words.
Respectively, a word when recognized by the subjects passes first through orthographic processing, then phonetic processing, and finally semantic processing. According to Craik and Lockhart, when a word has gone through these three levels of encoding it is more likely to be stored and later retrieved than a word that has only gone through the first level of processing. This lab tested a subject's ability to store, or hold learned information for later use, based on the three levels the word was possibly processed on. There were thirty-six words presented to each subject in the Psyk. Trek experiment. These words were processed by the subject on one of the three levels included in the Depth of Processing Model.
When the subjects were asked to retrieve the words presented to them they were first asked to bring up information from their memory with out clues, this process is known as free recall. Next, the subjects were asked to retrieve words based on clues, or the process of recognition. The data collected from the Memory lab then examined the subject's ability to retrieve 3 stored knowledge according to the level on which it was encoded and could be remembered. There were several hypotheses drawn prior to the conduction of the memory lab.
The first was that subjects would be more likely to remember words by recognition, rather than free recall. The second hypothesis was that words stored on a higher level of encoding would more likely be retrieved than those words stored on a lower level. The third hypothesis maintained that since there was a more in depth process of retrieval regarding higher levels of encoding, such as semantic processing, the time in seconds for the retrieval of a particular word would more likely be longer in duration than that of a word retrieved on a lower level, such as orthographic processing. Method Participants: The subjects in this experiment were collegiate students at the University of Connecticut. There where 8 male and 12 female participated in the experiment, there were a total of 20 students ranging from 18 to 20 years of age. Design: The Independent variables in this experiment were structural encoding, phonemic encoding, and semantic encoding.
The dependant variables were the number of words retrieved in recall and recognition, and the time in seconds it took to retrieve specific encoding levels from memory. Materials: The materials used in the Memory Lab were Psyk. Trek computer software, the memory program, on the Macintosh pc. Procedure: The procedure for this experiment began when the subject opened Psyk. Trek, and turned the cube in the center of the screen to "simulations." The subject then clicked on the cube and chose the icon "Memory Process." Next, a brief explanation of the trials 4 appeared on the screen. The subject proceeded to read this description and instructions.
To begin trials, the subject clicked on the red box icon on the screen. The subject answered questions that required the encoding of the word on the screen structurally, phonemically, or semantically, by clicking on the yes or no icon on the screen. The subject underwent thirty-six trials of such encoding questions. Following these trials, the subject was asked to recall as many words possible from the previous exercise.
After each word the subject pressed enter to verify the word entered on the screen. After the subject finished typing all of the words recalled, they pressed the done icon on the screen. Next, the subject underwent the recognition portion of the experiment. Here, the subject clicked on words they believed to previously have seen in the first part of the experiment when listed with two other words. The subject completed these thirty-six trials, and then saw the results regarding their performance provided by the computer. The data was represented in four different graphs; comparison of subjects' recall and recognition scores, serial position effects, response time for semantic, phonetic, and structural processing, and recognition by type of encoding.
The subject then read definitions regarding the independent variables of the experiment, and then answered a few quiz questions. Since there was no print option for a subject's individual data the subject then copied the information presented to them. This information was then used to calculate class means in order to further analyze the results. Results All of the three hypotheses under the limitations of this specific experiment were proven correct. The first hypothesis stated was that it was more likely for the subject to 5 retrieve words when prompted by clues rather than relying solely on free recall. The data recorded by class means supports this assumption, because the average number of words retrieved by free recall was 9.
40, whereas the number of words retrieved when prompted was 30. 06. The second hypothesis was in support of Craik and Lockhart's Depth of Processing Model. The class means for the different encoding processes showed that a word requiring a higher level of processing also had a higher level of retrieval.
The structural process, the lowest form of processing, received the lowest score for words retrieved from memory. The average number of structural words that were encoded was 8. 4. The phonetic received the next highest score for average number of words encoded under the auditory category. This level received a mean of 10. 8 words that were retrieved from memory.
Which was also the same rate in which the semantics process was received at. The last hypothesis discussed the fact that since a higher level of processing is achieved through, at first, lower levels of processing, it will take a longer duration in seconds to process a higher level of thought than a lower one. This hypothesis was supported by the class data since lower level structural processing had a retrieval time of 1. 22 seconds, whereas respectively, phonetic and semantic processing, higher processing levels had second durations of 1.
48 and 1. 72 seconds. Also, the independent variables represented levels of encoding in which supported the third hypothesis on how they effect and manipulate the dependent variable. In each case, from a lower to higher level of encoding, the dependant variables reflected the time of reaction to the encoded words recognized. The dependant variable increased when manipulated by an increased level of processing. 6 Discussion With the results found in the Memory Lab experiment, it can be concluded from group data that the higher the level of encoding involved in the processing of a word, the more likely it can be retrieved successfully from memory, and the higher the level of encoding involved in storage also increases the time it takes to retrieve such a word from memory.
The first hypothesis shows that words are three times more likely to be retrieved based on recognition rather than free recall. This is significant, because it demonstrates an increased response in a person's knowledge when prompted, as opposed to not being given a specific context in which a subject may work. Therefore, association by any of the three encoding processes allows a subject to better demonstrate knowledge. The second hypothesis stated that a word when stored in memory under a higher level of processing was recalled and recognized more often than a word stored under lower levels of encoding. The second hypothesis is significant, because it proposes methods to allow people to retain more information.
This is important, because everyone has to memorize information. This experiment concludes that information when processed with semantic meaning is more likely to be retrieved, therefore, the most effective way for a subject to study would be to learn by associating a word within a specific context in which the word held significant meaning. Structural and phonetic methods of encoding like trying to memorize vocabulary in the order it appeared on a page, or by sounding the word out are respectively less effective in retrieval. Finally, the third hypothesis shows how the independent variable affects the dependant variables. This is significant to learning about memory processes, because the reaction time to words encoded on higher levels 7 respectively takes a longer time period to process information regarding those words. Since words processed on a semantic level must first be processed on structural then phonetic levels, the time it takes to retrieve these words is dependent on the level at which a word is stored.
Two ways in which this experiment could be improved include results regarding effects of serial positioning, internal validity, and external validity. Serial positioning wasn't a valid method of evaluation in this experiment, because it only pertains to free recall. Data regarding free recall with serial positioning wasn't recorded, and could have been used as important information to evaluate the effects of serial positioning on free recall. The external validity of this experiment could have been improved by using a larger sample size of subjects to obtain a more accurate table of data rather than basing an experiment on selected students with relatively the same ability, in the same college course, all participating for the same reasons.
The internal validity in this experiment could have also been improved if serial positioning was included in the experimental data. Since serial positioning was important in the conduction of this lab, its lack of final representation was also a lack of internal validity. To further research on the topic of memory, not only can these faults be addressed, but also if words were repeated under the same encoding processes and then tested more than once, a better understanding of how effective each process is may be reached. Also different sections could be added to the experiment that called for the retrieval of a word on another level from which it was stored and on the same level it was stored. The reaction times it takes to process these words on different levels could prove as interesting data, either to support that a word is 8 retrieved at the same time for the level on which it was stored, or that a word's retrieval depends on its context, not the level at which it was stored..