Memory Ronald T. Kellogg\'s Working

Memory

Ronald T. Kellogg's "Working Memory Components in Written Sentence Generation": A Review and Further Research Inspired by the Study

In his article, "Working Memory Components in Written Sentence Generation," Ronald T. Kellogg used quantitative research in order to how working memory is impeded during distraction. Kellogg begins his article with a literature review detailing the types of research that psychologists have completed regarding memory in the past. He cites Baddely's 1986 work, which created the Baddeley model, a "phonological loop for storing and rehearsing verbal representations, a visuospatial sketchpad for visual object representation and their locations, and a central executive for attentional and supervisory functions" (341). In addition to this model, Kellogg also more modern research by Jonides and Smith that suggests the "visual, spatial, and possibly semantic stores are dissociable from the verbal store" (341). By considering both this old and new research concerning working memory, Kellogg designed his own research in order to increase psychologists' understanding of working memory.

Purpose

Kellogg designed his study in order to increase others' understanding of working memory. Specifically, Kellogg was interested in understanding if "planning conceptual representations" and "linguistically encoding these into words and sentences" depends on working memory (341). According to Kellogg, putting together a sentence, or sentence generation, requires "planning conceptual content," or deciding what one wants to say in writing, and "linguistically encoding it into a grammatical string of words," or placing those ideas and concepts into a well-formed, grammatically correct sentence (341). If one or either depended on working memory, a second purpose of the study was to review the strength of this dependence.

Methods

College students in a General Psychology class were chosen as the subjects to be tested. The students were assembled, given keyboards, and then given a visual prompt of two words, nouns. Subjects were then to write a "meaningful sentence" using the two nouns (344). At the same time, students were told to complete a "memory task," such as remembering certain digits. First, students' typing speed was assessed via trials. Next, students were given directions, followed by the two visual prompts, and then time to type their sentences and submit them to a computer. Finally, students were asked to complete the memory task they had been assigned correctly and were given feedback concerning whether or not they had made the correct response. Some students were asked to write sentences using nouns that were related, while others were asked to write sentences using nouns that were unrelated. In one group, students were asked to produce complex sentences; while in the other students were told to write simple sentences (344).

IV. Findings

In accordance with the methods above, the researchers derived results concerning, initiation time, sentence length and typing time, grammatical and spelling errors, and concurrent task performance. Only those students who were asked to remember six digits produced shorter sentences, while the memory tasks did not affect the length of any other students' sentences. Similarly, spelling and grammar were not affected by the memory tasks. Students writing using unrelated nouns took more time both to initiate and type their sentences then students using related nouns (347-350).

Significance

According to the author, the article was most significant finding of this study was that "neither a moderate load on verbal working memory nor a load on the visual and spatial components of working memory reliably impeded written sentence generation, but a heavy load on verbal working memory did so. Because the results were the same for all groups -- both simple and complex and both unrelated and related nouns -- the author safely assumes that the "heavy load" on working memory disrupts sentence generating and writing skills (350).

Experiment II Purpose

Because it is essential to functioning both psychologically and physically, working memory is an important topic to research and study. Though Kellogg developed a study with significant findings, working memory can effect more than just sentence generation. Because "heavy loads" on working memory do affect sentence generation, it is reasonable to assume that they might also affect simple mathematical processes. Do "heavy loads" placed on working memory affect just sentence generation or writing skills, or do they also affect left-brained skills like simple math calculation? Because of the effect on right-brained skills, the assumed hypothesis for this experiment is that if college students are given a "heavy load" on working memory, they will take longer to complete simple math skills. In fact, one would hypothesize that the "heavy load" on a working memory would impact mathematical skills even more than sentence generation skills because the subjects are being asked to remember a large number and then work with numbers. For this experiment, data using both a "heavy load" consisting of a numeric memory trial will be collected, though further research should consider performing the same task with a memory load that is not numerical and comparing the data.

Experiment II Method second group of college students will be selected for this experiment. Students will be selected from non-mathematical disciplines and will have had average mathematical scores on SAT and ACT tests. The students will then be given a series of computer-based mathematical timed tests in order to assess their ability to perform simple math skills. The simple skills will encompass making change, calculating mileage, and completing simple multiplication problems. Students will be given 10 second to initiate the task and 40 second to complete it. Before the testing begins, the students will run 10 trials in order to determine the rate at which they can complete the simple mathematical skills with no load on their working memories.

After the 10 trials, students will be given a visual memory trial on their screen just prior to the mathematical task. As Kellogg previously discerned that a six-digit memory task was considered a "heavy load," students will be asked to remember a series of six digits. They will have 7 seconds to memorize the digits, before the mathematical trials will begin. Students will complete the mathematical trials, after which they will see a series of six dots. Students will type the digits from the memory trial, replacing each of the dots. They will then press enter to submit the digits. In an attempt to correct a problem in experiment design from Kellogg's research, the students will not be given feedback concerning whether their answers were right or wrong as this may distract the students from using well-portioned time during the rest of the trials.