ONLY for One Touch! Psych Critical Evaluation Paper
What Predicts Skill in Lecture Note Taking?
Stephen T. Peverly, Vivek Ramaswamy, Cindy Brown, James Sumowski, and
Moona Alidoost Columbia University
Joanna Garner Cognitive Learning Centers
Despite the importance of good lecture notes to test performance, very little is known about the cognitive processes that underlie effective lecture note taking. The primary purpose of the 2 studies reported (a pilot study and Study 1) was to investigate 3 processes hypothesized to be significantly related to quality of notes: transcription fluency, verbal working memory, and the ability to identify main ideas. A 2nd purpose was to replicate the findings from previous research that notes and verbal working memory were significantly related to test performance. Results indicated that transcription fluency was the only predictor of quality of notes and that quality of notes was the only significant predictor of test performance. The findings on transcription fluency extend those of the children’s writing literature to indicate that transcription fluency is related to a variety of writing outcomes and suggest that interven- tions directed at transcription fluency may enhance lecture note taking.
Keywords: lecture note taking, study skills, transcription speed, cognitive processing, expertise
Contemporary views of expertise and cognitive processing sug- gest that performing a skill well usually depends on the parallel execution of two or more skill-specific processes within a limited- capacity working memory system.1 First, domain- or skill-specific basic skills (e.g., the processes that underlie word recognition) must be executed with an acceptable degree of fluency or auto- maticity, so that most, if not all, of the available space in working memory can be used for the application of the higher level cog- nitive skills (e.g., language ability) needed to produce successful outcomes (e.g., good comprehension). If basic skills are not au- tomatized, the application of higher level cognitive skills can be attenuated and prevent students from achieving their educational goal (e.g., Anderson, 1990; Baddeley, 1998, 2000; Ericsson & Kintsch, 1995; Kintsch, 1998; Perfetti, 1986; Schneider & Shiffrin, 1977; Shiffrin & Schneider, 1977). Second, individual differences in the capacity of working memory can lead to differences in the efficient execution of processes in working memory, which also can lead to differences in skill outcomes (Baddeley, 2001; Just & Carpenter, 1992; Swanson & Siegel, 2001). In other words, greater capacity in working memory enables greater efficiency in the processing and monitoring of higher order information (e.g., ap- plication of the knowledge of the language to interpret words). Finally, individual differences in higher level cognitive resources also can account for individual differences in task outcomes. In reading, for example, if word recognition (a basic skill) is autom- atized, individual differences in reading comprehension are highly correlated with language ability (Rayner, Foorman, Perfetti, Pe-
setsky, & Seidenberg, 2001; Vellutino, Fletcher, Snowling, & Scanlon, 2004).
Although we know a great deal about the development of expertise in a number of domains (Anderson, 1982; Chi, Glaser, & Farr, 1988), we do not know much about the cognitive skills that underlie expertise in lecture note taking. Past elementary school, most teachers communicate information through lecture (Putnam, Deshler, & Schumaker, 1993), and lecture notes, a cryptic written record of important information presented in class (Piolat, Olive, & Kellogg, 2005), are an important part of academic studying for adolescents and young adults (Thomas, Iventosch, & Rohwer, 1987). Most college students, for example, rate lecture note taking as an important educational activity (Dunkel & Davy, 1989), and most take notes in classes (approximately 98%; Brobst, 1996; Palmatier & Bennett, 1974). In addition, research has shown that recording (encoding) and reviewing notes from classes is related to good test performance (Bretzing & Kulhavy, 1981; Fisher & Harris, 1973; Kiewra, 1985; Kiewra et al., 1991; Kiewra & Fletcher, 1984; Peverly, Brobst, Graham, & Shaw, 2003; Rickards & Friedman, 1978; Titsworth & Kiewra, 2004).
Our and others’ analyses of note taking (Kiewra & Benton, 1988; Kiewra, Benton, & Lewis, 1987; Kobayashi, 2005; Peverly, 2006; Piolat et al., 2005) suggest that it is a difficult and cogni- tively demanding skill—students must hold lecture information in verbal working memory (VWM); select, construct, and/or trans- form important thematic units before the information in working memory is forgotten; quickly transcribe (via writing or typing) the information held in working memory, again before the information is forgotten; and maintain the continuity of the lecture (which also
1 Working memory is defined by most as storage and processing (e.g., Baddeley, 2001). There are a least four different categories of working memory theories, and each proposes a different explanation for individual differences in working memory. See Miyake and Shah (1999) and Peverly (2006) as well as the General Discussion of this article.
Stephen T. Peverly, Vivek Ramaswamy, Cindy Brown, James Sumowski, and Moona Alidoost, Teachers College, Columbia University; Joanna Garner, who is now at the Department of Applied Psychology, The Pennsylvania State University—Berks.
Correspondence concerning this article should be addressed to Stephen T. Peverly, Teachers College, Columbia University, Box 120, 525 West 120th Street, New York, NY 10027. E-mail: stp4@columbia.edu
Journal of Educational Psychology Copyright 2007 by the American Psychological Association 2007, Vol. 99, No. 1, 167–180 0022-0663/07/$12.00 DOI: 10.1037/0022-0663.99.1.167
167
consumes working memory resources). Thus, expertise in note taking may be related to three variables: transcription fluency, working memory, and the higher level processes needed to identify important information in lecture. Hypothetically, inadequate lec- ture notes could result from a breakdown in any one of these variables. For example, because of the substantial cognitive load typically present during lecture (Piolat et al., 2005), slow transcrip- tion speed could strain the capacity limitations of working memory and cause students to forget some of the information in working memory (through decay or interference) and lose continuity of the lecture.
Transcription Fluency
We were not able to find any research on the relationship of transcription fluency, the rate of written word production (Ransdell & Levy, 1996; Ransdell, Levy, & Kellogg, 2002), to the quantity or quality of lecture notes. However, there is indirect evidence for the importance of transcription fluency to writing outcomes among children and adults. Research on writing among elementary and middle school students suggests that (a) students’ transcription fluency (typically measured as the number of letters students can print or write in cursive in a minute) is related to the quality of their written compositions (Graham, Berninger, Abbott, Abbott, & Whitaker, 1997; Jones & Christensen, 1999) and (b) instruction in transcription fluency (how to properly form letters) in elementary school is related to improvement in the amount (Berninger et al., 1997; Graham, Harris, & Fink, 2000; Jones & Christensen, 1999) and quality of written products (Jones & Christensen, 1999).
Among adults, research has not typically focused on the rela- tionship of individual differences in transcription fluency and writing outcomes. Rather, research has focused on the effects of experimental manipulations of transcription fluency (e.g., writing in the way that one normally would vs. writing in uppercase cursive), the influence of the difficulty of a concurrent task (e.g., writing vs. copying an essay) on writing speed, and the ability to monitor processing in working memory (e.g., metacognitive pro- cesses, such as planning, revising). Results indicate that adults’ transcription fluency is faster under normal than under modified conditions and that slower transcription fluency is associated with poorer monitoring of processing in working memory and more errors in the recall of information from working memory (J. S. Brown, McDonald, Brown, & Carr, 1988; Olive & Kellogg, 2002).
We know of two studies that have evaluated the relationship of individual differences among adults in transcription fluency to essay quality. Connelly, Dockrell, and Barnett (2005) evaluated the relationship of transcription fluency to the quality of under- graduate students’ essays under two conditions— unpressurized and pressurized. In both conditions, all the students in a 2nd-year psychology class (n � 22) had an hour to write an essay. In the former condition, all of the students wrote a practice essay in preparation for a final exam. In the latter, the students wrote an essay as part of an end-of-semester examination. It was hypothe- sized that the pressure of a real examination would increase students’ cognitive load and thus create a stronger relationship between transcription fluency and exam performance. Transcrip- tion fluency was measured by a modification of the alphabet task, a measure of handwriting fluency (Berninger, Mizokawa, & Bragg, 1991). In this modification, students are told to write the
alphabet in lowercase letters as many times as they can in 1 min. Students’ essays were scored in three different ways: scores given to the essays by course tutors, number of words written (for the entire essay as well as for the introduction, main body, and conclusion), and rubric assessment scores (the rubric “assessed students’ skill at sectioning the essay clearly, ordering ideas, linking ideas, showing sufficient support and expansion of ideas and showing a sufficient sense of audience”; Connelly et al., 2005, p. 100). Results indicated that there were no significant correla- tions between handwriting fluency and any of the essay scores in the unpressurized condition. In the pressurized condition, however, transcription fluency correlated positively and significantly with tutors’ marks, overall number of words written, and the overall rubric score. These data, along with data from research on the experimental manipulation of transcription fluency, suggest that transcription fluency is related to working memory and to writing quantity and quality, especially in situations in which there is a substantial degree of cognitive load.
Connelly, Campbell, MacLean, and Barnes (2006) evaluated the effects of lower level writing skills (transcription fluency as mea- sured by the alphabet task and spelling skill), higher level writing skills (e.g., vocabulary; organization, unity and coherence), and other cognitive variables (e.g., VWM) on essay writing among three groups: college students with dyslexia, an age-matched group of college students without dyslexia, and a spelling skill control group (ages 11 to 31; an average age of 18) whose spelling skills matched those of the dyslexic group. For our purposes, the results indicated that the essay writing skills of the nondyslexic college students were superior to those of the other two groups, who were not different from each other, and that transcription fluency, as measured by the alphabet task, was related to essay quality for the dyslexic and nondyslexic college students but not the spelling skill control group.
In the experiments reported in this article, we use two fluency tasks to evaluate which might correlate better with lecture notes: the alphabet task and the Writing Fluency subtest of the Woodcock–Johnson Psychoeducational Battery—Revised (Tests of Achievement, Form A; Woodcock & Johnson, 1989). Both have been used in research to evaluate the transcription fluency of children and adults. We included both in an attempt to isolate the factors related to transcription fluency. The alphabet task allowed us to measure students’ speed of forming the units (letters) that are the foundation of words unencumbered by other skills that might affect the speed of writing words (e.g., knowledge of orthography or syntax). The Writing Fluency subtest measures the speed of writing short sentences of the type students might use in taking notes.
Working Memory
Research indicates that interindividual differences in working memory are positively and strongly related to a wide variety of skills (e.g., reading and writing) and abilities (e.g., verbal ability; Baddeley, 2001; A. D. Baddeley, personal communication, De- cember 9, 2004; Bayliss, Jarrold, Gunn, & Baddeley, 2003; Dane- man & Carpenter, 1983; Just & Carpenter, 1992; Kellogg, 2001, 2004; Swanson & Berninger, 1996; Swanson & Siegel, 2001) and that taking notes from lectures is very demanding of working memory resources (Piolat et al., 2005). The relatively small
168 PEVERLY ET AL.
amount of research on the relationship of VWM to the quantity and quality of notes has produced mixed results, however. Kiewra and Benton (1988; Kiewra et al., 1987) and McIntyre (1992) found that working memory was related to the quantity and quality of notes, but Cohn, Cohn, and Bradley (1995) found that it was not.2
The lack of consistent outcomes between working memory and notes may be due to differences among studies in the measures used to evaluate it. Kiewra and Benton (1988; Kiewra et al., 1987) and McIntyre (1992) used tasks that required participants either to unscramble randomly ordered words to make a sentence (six sentences in total) or to arrange randomly ordered sentences to make a coherent paragraph. These tasks are different from the complex span tests typically used to assess VWM.3 For example, one commonly used complex span task is Daneman and Carpen- ter’s (1980) reading span test, which requires participants to read a set of unrelated sentences (two to five) one at a time. As soon as they have finished reading one sentence, the next sentence is presented, and the procedure is repeated. Once the participants come to the end of the set and all of the sentences have been removed, they are asked to remember the last word of each sentence. In the tasks used by Kiewra and Benton (1988; Kiewra et al., 1987) and McIntyre (1992), participants had all of the materials in full view during the entire task. Because these tasks do not require participants to remember and process information in the same way as the complex span tasks, they may not adequately measure either span or processing as they are typically conceived in the working memory literature.
Cohn et al. (1995), who did not find a significant relationship between working memory and notes, used three of the working memory tasks used by Turner and Engle (1989) in their research on working memory: operation–word, sentence–word, and word span. All are complex span tests of the type used by Daneman and Carpenter (1980). Sentence–word, for example, is like reading span except that participants must also judge whether the sentences make sense. In the two experiments reported in this article, we used Daneman and Carpenter’s (1980) listening span task. It has the advantage of being similar to the reading span task (with the exception that participants listen to sentences rather than read them and make judgments about the meaningfulness of the sentences) and to the other complex span tasks commonly used in research on working memory (e.g., Daneman & Carpenter, 1980, 1983; Engle, 2001, 2002; Swanson & Siegel, 2001). Also, from an ecological perspective, it is a better match to a listening-based task such as taking lecture notes than is reading span.
Identification of Main Ideas
A well-organized macrostructure—that is, a summary of the main themes and ideas in spoken or written discourse—is crucial to students’ demonstrations of learning and remembering of what they have heard or read (Kintsch, 1998). In the context of note taking, students favor important over less important propositions in notes (Bretzing & Kulhavy, 1981; Kiewra & Fletcher, 1984; Rickards & Friedman, 1978; Wade & Trathen, 1989), and the amount and quality of information in notes are related to test performance (Cohn et al. 1995; Kiewra & Benton, 1988; Peverly et al., 2003). Peverly et al. (2003), for example, found that the number of macropropositions in text notes (the logical or rhetorical relationships among propositions that describe the thematic struc-
ture of discourse) was directly related to measures of students’ learning from text.
Finding a measure of students’ ability to identify main ideas that is correlated with notes is not straightforward, however. Kiewra and Benton (1988) and Kiewra et al. (1987) found that American College Test Comprehension and English scores and grade point average (GPA), measures that one might assume would be related to the ability to identify main ideas, were not significantly corre- lated with the contents of notes. In addition, Peverly, Brobst, Shaw, and Graham (1998) found that vocabulary scores were not significantly correlated with notes. Vocabulary correlates highly with reading comprehension (Kintsch, 1998) and verbal IQ (Satt- ler, 2001), and the latter correlates highly with text comprehension once word recognition is automatized (Rayner et al., 2001; Vellu- tino et al., 2004).
Kintsch (1998) argued that two of the more important skills related to successful text comprehension are the deletion of unim- portant information (trivia and redundancy) and the identification or construction of main ideas. Given the lack of success with other measures of comprehension and verbal skill, we constructed a task to measure students’ ability to differentiate between important and unimportant information more directly. Students were asked to read a four-page, double-spaced text on the rise and the fall of the Roman empire and to label each of 20 statements from the text as a main idea or a detail. This task is described in more detail in the Method section of the pilot study.
2 Other studies have been cited in the literature in support of the relationship between VWM and lecture note taking (e.g., DiVesta & Gray, 1973; Peters, 1972). However, from our vantage point, their data are difficult to interpret. First, DiVesta and Gray (1973) did not provide much of a description of their task other than to say that they used “a memory span test patterned after Peterson and Peterson’s (1959) short-term memory task” (p. 281). Peterson and Peterson (1959) gave participants consonant– consonant– consonant strings (e.g., DNT) and, to prevent rehearsal after presentation, required participants to count backward by 3s from a number they were given. The researchers varied the retention interval (how much time participants spent counting backward) before participants were asked to recall the string of letters. Their purpose was to evaluate the rate of decay in short-term memory not short-term memory itself. Also, DiVesta and Gray generated 64 correlations between their measure of short-term mem- ory and other variables in the experiment. Only 2 were significant. They stated, “Because of the number of correlations calculated these may have occurred by chance, and any conclusions can only be suggestive” (p. 284). In addition, Peters (1972) did not use a measure of short-term or working memory as they are typically defined (and did not use the words short-term or working memory to describe his task or results). He created what he called a learning efficiency measure. It was composed of two lists of 20 items each. Each item consisted of a social psychological term and a definition. One list was recorded at 130 words per minute and the other at 192 words per minute. Each list was presented followed by a test during which participants heard the definition and had to fill in the term associated with it. The difference between participants’ performance on Lists A and B was used as a measure of their learning efficiency. Although one can assume that working memory was involved in this task, other factors also must have played a role (e.g., long-term memory).
3 Some authors have referred to these tasks as information processing tasks (e.g., McIntyre, 1992), and others have referred to them as both working memory and information processing tasks (e.g., Kiewra et al., 1987).
169SKILL IN LECTURE NOTE TAKING
Purpose
We conducted a pilot study to replicate the finding that notes are a strong predictor of test performance but, most important, to evaluate the relative contributions of transcription fluency, VWM, and the ability to identify main ideas to the quantity (the number of topics students mentioned in their notes) and quality (how well students explained each topic) of students’ lecture notes. Relative to the latter, we also included a measure of spelling skill to evaluate whether it is related to the quantity or quality of students’ notes, given the findings that skill in spelling is related to tran- scription fluency in younger elementary grade students (Graham et al., 1997) and that instruction in spelling transfers to improvement in transcription fluency (Berninger et al., 1998).
Given the substantial amount of evidence on the relationship between notes and test performance and the finding by Cohn et al. (1995) that notes and VWM were related to performance in an economics course, we hypothesized that both would independently predict test performance. In addition, given the findings from research on individual differences in writing speed and experimen- tal manipulations of writing speed on measures of quantity and quality of essays among children and adults, we hypothesized that transcription fluency (including spelling) would be positively re- lated to the quality and quantity of notes. Also, despite the am- biguous relationship between VWM and the quantity and quality of notes, we predicted that VWM would account for a significant portion of the variance in the quantity and quality of notes inde- pendent of that accounted for by transcription fluency, given its strong relationship to other verbally based skills, such as reading and writing. In addition, because the ability to identify main ideas is strongly related to reading comprehension (Kintsch, 1998) and studying (A. L. Brown & Day, 1983; A. L. Brown, Day, & Jones, 1983), we hypothesized that it would be related to the quantity and quality of lecture notes. The relationships evaluated in the pilot study are summarized in Figure 1.
Pilot Study
Method
Participants
Participants were undergraduate students (N � 85) in an introductory psychology course at a large university in the northeastern United States
who participated for course credit. Their mean age was 20.38 years (SD � 2.47), 75.3% were women, 65.9% spoke English as their first language, and 30.6% were psychology majors (74% reported that they had taken two or fewer college psychology courses). The race/ethnicity of the sample was diverse: White (42.4%), African American (5.9%), Asian (14.1%), Latino/a (16.5%), Native American (1.2%), and other (16.5%).
Materials and Scoring
The materials consisted of the lecture video, written summary, two measures of transcription fluency (the alphabet task and the Writing Fluency subtest of the Woodcock–Johnson Psychoeducational Battery— Revised; Woodcock & Johnson, 1989), a spelling test, the listening span task (VWM), and the main idea differentiation task. All measures were group administered. Interrater agreement in scoring (agreement/ agreement � disagreement � 100%) was established for all measures. Twenty protocols (approximately 25%) were randomly chosen, and two graduate students independently scored all of the measures in each partic- ipant’s protocol. Disagreements were settled by consensus.
Lecture
The lecture and the method used to score students’ lecture notes were taken from Brobst (1996). The videotaped lecture was 20 min long and summarized basic concepts and research in the psychology of problem solving. The lecture was read from a prepared text by Stephen T. Peverly. Participants were given two sheets of blank paper and told to take notes. They also were informed that they would be allowed 10 min to study their notes in preparation for an essay test sometime later in the study.
The content of the lecture was adapted from a chapter by Voss (1989) titled “Problem Solving and the Educational Process.” The lecture con- sisted of six themes (e.g., functions of problem solving in education), some of which were subdivided into separate content areas. There was a total of 15 content areas. The structure and content of the essay are detailed in the Appendix.
Participants’ notes were scored for quantity and quality. Quantity scores reflected the number of topics students mentioned in their notes. Students’ quantity scores could range from 0 to 15. Quality scores reflected the rating (0 –3) given to each of the 15 items mentioned. A rating of 0 was given for incorrect or missing information, a rating of 1 if a topic was mentioned but not elaborated, a rating of 2 for an incomplete explanation, and a rating of 3 for a complete explanation. Quality scores could range from 0 to 45. The quality ratings given to each of the 15 topics were item specific and specified in a manual created by Brobst (1996). Take, for example, Content 1 in the Appendix, which is important to the subareas of educational theory and classroom practice. A participant would be given 1 point for each concept mentioned. If a participant wrote, “Problem solving is a cognitive activity,” the statement would receive a score of 1. If a participant wrote, “Problem solving is a cognitive activity that is important to educational theory and classroom practice,” the statement would receive a score of 3. Interrater agreement for the randomly chosen protocols, collapsed across quantity and quality scores, was .91.
Written Summary
Participants were instructed to write an organized summary of the videotaped lecture without referring to their notes. They were allowed 10 min and given two sides of one sheet of paper for this task. The same method and criteria used for scoring notes were used for scoring essays (e.g., students’ quantity scores could range from 0 to 15, and their quality scores could range from 0 to 45). Interrater agreement was. 95.
Transcription Fluency
The alphabet task. This task is based on one used by Berninger et al. (1991) that asked children to write as many letters of the alphabet as they
VWM
Notes
Transcription Fluency
-Letter Fluency Test Performance-Compositional
Fluency
Identification of Main Ideas
Spelling
Figure 1. Pilot study: model of the relationship of transcription fluency, verbal working memory (VWM), and main idea identification to notes and the relationship of notes to test performance.
170 PEVERLY ET AL.
could in 30 s (hereafter referred to as letter fluency). In this study, participants were instructed to write the alphabet horizontally in capital letters on a blank sheet, starting with A. Once finished, they were to begin the alphabet again in lowercase letters and continue to alternate between lowercase and uppercase letters until the time expired. One point was awarded for each recognizable letter, and the points were summated to calculate participants’ total scores. Interrater agreement across 20 ran- domly chosen protocols was 1.00.
Writing fluency. Participants were group administered the Writing Fluency subtest of the Woodcock–Johnson Psychoeducational Battery— Revised (Woodcock & Johnson, 1989), a test of the ability to construct and transcribe simple sentences quickly (hereafter referred to as compositional fluency). In this subtest, participants were shown sets of three words accompanied by a picture stimulus. They had to write complete, semanti- cally and syntactically appropriate sentences that related to the picture and included all three words (none of the words could be changed in any way). The number of sentences completed during the 7-min time limit was summated to yield each participant’s total score out of a possible 40 points. Each sentence received a score of 1 or 0. A score of 1 was given if the sentence met all of the criteria mentioned in the previous sentences. Otherwise a score of 0 was given. The test–retest reliability of this subtest is .77, with a standard error of measurement of 7.1 for the 18-year-old age group (the closest age group to the participants in the study). Across 20 randomly chosen protocols, interrater agreement was .93.
Spelling
Participants’ spelling skills were assessed with the Spelling subtest of the Wide Range Achievement Test—Third Edition (Wilkinson, 1993). The 40 spelling words contained in the Blue Form of the subtest were dictated aloud and written by the participants in their test packet. There is no specified time limit for this test. (The other section of the Spelling subtest, Name/Letter Writing, was not administered.) One point was given for each word spelled correctly, and the points were summated for each partici- pant’s total score out of a possible 40 points. As reported in the test manual, the coefficient alpha is .93 for the 20 –24-year-old age group. The test– retest reliability, corrected for attenuation, is also .93. The interrater agree- ment for this measure was 1.00.
VWM (Listening Span)
The measure used to assess participants’ auditory VWM was the listen- ing span test (Daneman & Carpenter, 1980, Study 2). Participants were presented via audiotape with 60 unrelated sentences composed of five levels of three sentence sets each. The first level consisted of three sets of 2 sentences each. The next consisted of three sets of 3 sentences, and so on until the last set, which consisted of three sets of 6 sentences each. As participants listened to each sentence, they had to determine whether each sentence made sense and circle “yes” or “no” in their test packet. After each sentence set was completed, a beep prompted the participants to recall and write down the last word of each sentence in that set. After 20 s, another beep sounded, signaling the beginning of the next sentence set.
The scoring of the listening span task followed the procedures laid out in Daneman and Carpenter (1980). Scores on this measure were based on the highest level (2– 6) at which participants remembered all of the words for at least one of the three sentence sets. That is, if a participant correctly recalled all of the final words for two or all three of the sentence sets at Level 4 but none at Level 5 or 6, his or her score would be 4. If a participant correctly recalled all of the words for only one set at Level 4, the score was the number of sentences in that set minus 0.5 (3.5). Scores could range from 1.5 to 6 in increments of 0.5. Interrater agreement was .94.
Main Idea–Detail Differentiation Task
The text for this task was taken from Peverly et al. (2003). Participants were presented with a passage of approximately 1,000 words (four double-
spaced pages) on the rise and fall of the ancient Roman empire (readability of Grade 13). The overall structure of the passage was primarily chrono- logical (ranging from B.C. to A.D.). The passage consisted of 10 cause– effect sequences (e.g., Rome’s strategic location along the Tiber River and seven hills helped it control commerce and trade, which resulted in a wealthy and dynamic city) and one collection (one listing of items; in our text it was a listing of the legacies of the Roman empire; Meyer, 1985; Meyer & Poon, 2000). The introductory paragraph provided a general introduction to the two themes of the passage: (a) Rome’s shaping of the ancient Mediterranean world, and (b) Rome’s legacies and contributions to contemporary Western society. All of the remaining paragraphs but the last one developed the first theme. The last paragraph developed the second theme. Information on the procedures used to verify the content and structure of the passages (e.g., what was a macroproposition and what was not) can be found in Peverly et al. (2003).
Along with the passage, participants were given 20 statements relating to the content of the essay. Ten of the statements were main ideas (e.g., “The beginning of Octavian’s reign marked the end of the Republic and the beginning of the Pax Romana”), and 10 were less important information or details (e.g., “The Etruscans built a center market place, the Forum, which ultimately became the seat of Roman government”). The order of the statements was randomized. Participants had 10 min to read the passage and answer the questions with the text in front of them (by circling M for main idea and D for detail at the end of each statement). The number of correct responses to the 20 items was summated to yield a total score for each participant. Interrater agreement across all 20 randomly chosen pro- tocols was 1.00.
Procedure
Potential participants were given a packet of materials, with a consent form describing the purpose (i.e., “You are invited to participate in an experiment designed to examine the skills related to taking lecture notes”) and the tasks and time involved in the study as a cover sheet. If they signed the consent form, they were asked to turn the page and complete a short demographics questionnaire. Subsequently, they were told that they were going to watch a 20-min videotape on the psychology of problem solving (Stephen T. Peverly read the lecture from a prepared text). Participants were told to take notes on two pieces of paper provided in the packet of materials. They were also told that they would have 10 min to study their notes sometime later in the study and that, because they would have only their notes to study from, it was important that their notes be as complete as possible. After the lecture was completed, the remaining tasks of the study were administered in the following order: letter fluency, spelling, VWM, 10-min study period, composition fluency, essay, and main idea task. The entire study took approximately 90 min.
Results
Although a path analysis is typically used to evaluate relation- ships of the type depicted in Figure 1, the sample was too small (Kline, 1998). Thus, the data from the pilot study were analyzed with regression analyses. In the first regression, recall quality was the dependent variable, and transcription fluency (letter fluency, composition fluency), spelling, VWM, notes’ quality, and identi- fication of main ideas were the independent variables. In the second set of regression analyses, quality of notes was the depen- dent variable, and all of the other variables, with the exception of recall quality, were the independent variables.
Table 1 contains the means and standard deviations for the independent and dependent variables. Table 2 contains the inter- correlations among the independent and dependent variables. The correlations in Table 2 indicate that notes’ quality was the only
171SKILL IN LECTURE NOTE TAKING
independent variable to correlate significantly with quality of written recall, main idea identification did not correlate signifi- cantly with any of the other variables (we eliminated this variable from all further analyses), and all of the remaining independent variables were significantly correlated with each other. The reader should note that notes’ quality and quantity were very highly correlated (.93), as were recall quality and quantity (.94; these are not reported in Table 2). We chose notes’ quality and recall quality to include in the regression equations because the quality scores had more variation and correlated a little better with the other variables. Finally, all variables were tested for normality and found to be within acceptable limits.
First, using a stepwise regression, we regressed recall quality on our measures of notes’ quality, transcription fluency (letter fluency and compositional fluency), spelling, and VWM to determine which of these variables was related to test performance. The regression equation was significant (tolerance and variance infla- tion factor values were within acceptable limits; R � .37, R2 � .14, Radjusted
2 � .13), F(5, 81) � 12.83, p � .001 (the effect size, with R2 used as an estimate of effect size, was small; Cohen, 1988). The only significant predictor was notes’ quality (�� .37, p � .001). See Table 3.
Next, using a stepwise regression, we regressed notes’ quality on transcription fluency (letter fluency and compositional fluency), spelling, and VWM to evaluate which variables were related to quality of notes. The regression equation was significant (tolerance and variance inflation factor values were within acceptable limits; R � .34, R2 � .11, Radjusted
2 � .10), F(4, 81) � 10.13, p � .002 (again, the effect size, according to Cohen, 1988, was small). The only significant predictor was letter fluency (b � .34, p � .002). See Table 4.
Discussion
Our hypothesis that notes and VWM would be related to test performance was only partially confirmed. Notes but not VWM were positively and significantly related to recall quality. The relationship of notes to test performance (recall quality) confirms previous findings (Bretzing & Kulhavy, 1981; Fisher & Harris, 1973; Kiewra, 1985; Kiewra et al., 1991; Kiewra & Fletcher, 1984; Peverly et al., 2003; Rickards & Friedman, 1978; Titsworth & Kiewra, 2004). The lack of a significant correlation between VWM and test performance, however, was not expected. Cohn et al. (1995), who used a VWM measure similar to ours, found a significant relationship between VWM and test performance, and VWM has been found to correlate significantly and positively with a variety of measures of verbal ability (e.g., SAT; Daneman & Hannon, 2001) and reading comprehension (e.g., Daneman & Carpenter, 1983; Swanson & Siegel, 2001), which, in turn, are usually correlated with test performance, although not as highly as notes (Kiewra & Benton, 1988; Kiewra et al., 1987). We explore some of the possible reasons for our finding in the General Discussion.
We also predicted that transcription fluency (letter fluency, compositional fluency) and VWM would predict notes’ quality. Again, our hypothesis received only moderate support; only tran- scription fluency, as represented by the letter fluency task, was significant. VWM was not a significant predictor. The latter may be explained by the pattern of correlations among the variables. VWM might have been too highly correlated with letter fluency to contribute a significant amount of additional variance in the re- gression equation. Finally, contrary to our prediction, our measure of main idea identification did not correlate significantly with any of the dependent variables or other independent variables. There may be two reasons for this. First, reading a four-page text and answering 20 questions in 10 min might have been too difficult. Although participants’ performance was significantly above chance, t(82) � 6.72, p � .000, the mean score of 11.80 was well short of perfect performance (20), no one attained a perfect score (the highest was 18), and there was relatively little variation in participants’ performance (SD � 2.43). The other reason might have been fatigue. Some of the participants complained that there were too many tasks for one session, and the main idea task was the last one the participants completed.
The finding that transcription fluency, operationalized as letter fluency, was related to notes’ quality extends findings from the children’s writing literature on the relationship of transcription fluency to the amount and quality of what children write and, along with J. S. Brown et al. (1988), Connelly et al. (2005), Connelly et al. (2006), and Olive and Kellogg (2002), provides evidence of the importance of transcription fluency to writing among adults both
Table 1 Pilot Study: Means and Standard Deviations
Statistic Recall Spelling Letter flu Notes Comp. fluency VWM Main idea
M 3.55 28.99 62.12 13.45 26.92 4.54 11.80 SD 2.45 3.54 22.81 5.55 5.74 1.29 2.43
Note. Letter flu � letter fluency; Comp. � composition; VWM � verbal working memory.
Table 2 Pilot Study: Intercorrelations Among the Independent and Dependent Variables
Variable 1 2 3 4 5 6 7
1. Recall — 2. Spelling .02 — 3. Letter flu .20 .36** — 4. Comp. flu .21 .49** .53** — 5. Notes .37** .22* .34** .29** — 6. VWM .11 .37** .44** .28** .29** — 7. Main idea .00 .02 .14 .06 .15 .05 —
Note. flu � fluency; Comp. � composition; VWM � verbal working memory. * p � .05. ** p � .01.
172 PEVERLY ET AL.
when they are generating ideas (e.g., writing essays) and when they are recording them (e.g., lecture notes). If this finding is replicated, it may have important implications for teaching and remediating lecture note taking.
Interpretation of precisely why letter fluency was related to notes’ quality is not straightforward, however. One possibility is that performance on this task is related to two variables: fine motor skills (planning and production of letter forms) and the speeded access to verbal codes (phonetic units associated with letters of the alphabet). Certainly, there is evidence in this study to support the latter. Letter fluency was correlated with all of the other indepen- dent variables, which are all verbally loaded, with the exception of the main idea task, which did not correlate with anything.
In addition, there is evidence to support the relationship of fine motor skills and the access of verbal codes to transcription fluency in the children’s writing literature (Abbott & Berninger, 1993; Berninger et al., 2006; Berninger & Hooper, in press; Berninger & Richards, 2002). Abbott and Berninger (1993), for example, in a cross-sectional study of the development of writing skill among children in the first through sixth grades, found that fine motor skill did not predict transcription fluency directly but was mediated by orthographic coding, which is highly verbally loaded. Relat- edly, Berninger et al. (2006), in another developmental study, found that graphomotor planning and orthographic coding were related to cursive writing in the third grade and that orthographic coding and executive planning were related to cursive writing in the fifth grade. Thus, at least in elementary and middle school children, both fine motor skill and verbal codes are implicated in transcription fluency, although the strength of the latter is greater than the former. In an effort to evaluate these relationships more thoroughly in Study 1, we added a task that required participants to write nonalphabetic, nonverbally loaded symbols quickly to
determine whether a task that relied more on fine-motor speed would be related to notes’ quality.
Finally, there were some problems with the sample. English was not the first language for about one third of the sample, and Dunkel, Mishra, and Berliner (1989) found that native speakers of English recalled more information from lectures than nonnative speakers. Thus, we evaluated the differences between native and nonnative English speakers on the independent and dependent variables included in the analyses. The means and standard devi- ations are in Table 5. Although none of the effect sizes was large, significant differences in favor of native English speakers were found on compositional fluency, F(1, 81) � 9.55, p � .003 (d � 0.11); VWM, F(1, 81) � 10.62, p � .002 (d � 0.12); and spelling, F(1, 81) � 4.47, p � .038 (d � 0.05). Also, approximately 31% of the participants were psychology majors. There were no signif- icant differences between psychology and nonpsychology majors on any of the variables, with the exception of letter fluency. For some curious reason, psychology majors (M � 70.88, SD � 24.29) wrote letters of the alphabet faster than nonpsychology majors (M � 58.19, SD � 21.17), F(1, 82) � 5.9, p � .017 (d � .07). In addition, knowledge of the topic of problem solving might have confounded the outcomes. Although only 9 participants (10.5%) indicated that they had taken a psychology course that covered the topic of problem solving, this might have been enough to adversely affect the outcomes. Unfortunately, however, our sample was not large enough to systematically evaluate the effects of first lan- guage or major on the study’s outcomes.
Study 1
The purpose of this study is to replicate and extend the results of our pilot study, especially those related to the second hypothesis,
Table 3 Pilot Study: Summary of the Regression Analysis Predicting Test Performance
Variable B SE B � Partial r Tolerance VIF
Spelling �0.11 0.09 �.17 �.10 .93 1.08 Letter flu 1.55 0.01 .01 .03 .89 1.13 Notes 0.16 0.05 .37**** .37 1.00 1.00 Comp. flu 5.90 0.06 .14 .07 .92 1.09 VWM 6.66 0.23 .00 �.02 .92 1.09
Note. R � .37, R2 � .14, Radjusted 2 � .13. VIF � variance inflation factor; flu � fluency; Comp. � composition;
VWM � verbal working memory. **** p � .001.
Table 4 Pilot Study: Summary of the Regression Analysis Predicting Notes’ Quality
Variable B SE B � Partial r Tolerance VIF
Spelling 0.16 0.20 .10 .18 .88 1.14 Letter flu 4.51 0.03 .18 .34*** 1.00 1.00 Comp. flu 0.11 0.13 .11 .15 .74 1.35 VWM 0.62 0.51 .15 .13 .17 1.23
Note. R � .40, R2 � .16, Radjusted 2 � .10. VIF � variance inflation factor; flu � fluency; Comp. � composition;
VWM � verbal working memory. *** p � .002.
173SKILL IN LECTURE NOTE TAKING
on the relationship of transcription fluency and VWM to notes’ quality, with a larger, more homogeneous sample of students (N � 151) that included very few nonnative speakers of English and very few psychology majors. We attempted to extend our results by adding a measure of graphomotor fluency, the Symbol Coding subtest of the Wechsler Adult Intelligence Test—Third Edition (WAIS–III; Wechsler, 1997). This test requires participants to copy nonlinguistic shapes (e.g., �) as fast as they can. We added it to evaluate whether a measure of fine motor speed, not con- founded by phonological knowledge, would be related to quality of notes. Also, we added a measure of verbal fluency to evaluate participants’ speed of semantic access. Participants were given two tasks loosely modeled on those in the NEPSY (Developmental Neuropsychological Assessment) (Korkman, Kirk, & Kemp, 1998)—they had 1 min to write as many words as they could think of for each of two letters (F and S) and two semantic categories (animals and foods). McCutchen, Covill, Hoyne, and Mildes (1994) found that better writers have faster and more accurate access to words in their mental lexicon and thus may generate more ideas than poorer writers. In Study 1, we evaluated whether speed of semantic access was positively related to notes’ quality.
In summary, in Study 1 we evaluated whether transcription fluency (letter fluency, compositional fluency, digit symbol), ver- bal fluency (phonetic and semantic), and VWM were related to quality of notes and again whether quality of notes was related to test performance. We did not include spelling or main idea iden- tification because they did not uniquely predict variance associated with notes or recall in the pilot study. The model evaluated in this study is presented in Figure 2.
Method
Participants
Participants were undergraduate students in an introductory psychology course at a large, public university in central Pennsylvania (N � 151) who participated for course credit. Their mean age was approximately 20.07 years (SD � 2.22), and 86% (n � 130) of the participants were women. The sample was very homogeneous. Over 90% of the participants de- scribed their ethnicity as White, and only 6 participants (4%) reported that they were nonnative English speakers. The participants had a limited background in psychology, as only 10 of them (7%) described themselves as psychology majors or minors, and only 9 participants (6%) reported having taken more than three college psychology courses.
Materials
All of the materials and the administration of the materials were the same as in the pilot study, except as noted. Also, all materials were scored
by three graduate students instead of two. One of the three trained raters, who was a rater in the pilot study, trained the other two. Thus, we calculated interrater agreements by comparing the trainer with each of the trainees on 25 protocols. Disagreements were settled by consensus.
Lecture Notes, Written Recall, Letter Fluency, and VWM
The range of interrater agreement was .94 to .95 for lecture notes (only notes’ quality was scored), .94 to .96 for the written recall, and .99 to 1.00 for letter fluency. Interrater agreement for VWM was 1.00.
Phonetic and Semantic Retrieval
The phonetic and semantic retrieval tasks were based on the Verbal Fluency subtest of the NEPSY (Korkman et al., 1998). These tasks assess individuals’ ability to fluently access words in memory on the basis of phonetic or semantic cues. For the two phonetic retrieval tasks, participants were given 1 min to write down as many words as they could that began with the letter S. The task was repeated with the letter F. For the two semantic retrieval tasks, participants were given 1 min each to write down as many words as possible that belonged to the categories animals and food and drink. The number of correct responses was evaluated on the basis of the scoring rules in the NEPSY manual (e.g., no repetitions, proper names, or different forms of the same word). The scores from the two phonetic retrieval tasks were combined, as were the scores from the two semantic retrieval tasks. Interrater agreement was .99 for phonetic retrieval and .99 for semantic retrieval.
Letter Fluency
VWM
Notes Quality
Comp.
tseTlobmyS PerformanceDigit
Semantic Fluency
Fluency
Phonetic Fluency
Figure 2. Experiment 1: model of the relationship of letter and compo- sition (Comp.) fluency, verbal working memory (VWM), and phonetic and semantic retrieval to notes and the relationship of notes to test performance.
Table 5 Pilot Study: Means and Standard Deviations for Native and Nonnative English Speakers
Participant
Notes Recall Comp. flu VWM Main idea Letter flu Spelling
M SD M SD M SD M SD M SD M SD M SD
Native English speaker (n � 56) 14.02 5.43 3.89 2.62 28.20 5.98 4.85 1.18 11.96 2.33 65.30 22.58 29.48 3.12 Nonnative English speaker (n � 27) 12.08 5.73 2.96 1.89 24.22 4.26 3.91 1.34 11.52 2.68 56.19 22.61 27.78 4.03
Note. Comp. � composition; flu � fluency; VWM � verbal working memory.
174 PEVERLY ET AL.
Writing (Compositional) Fluency
Participants were administered the Writing Fluency subtest of the Woodcock–Johnson III (Tests of Achievement, Form A; Woodcock, McGrew, & Mather, 2001), not the Writing Fluency subtest of the Woodcock–Johnson Psychoeducational Battery—Revised, which was used in the pilot study. In the interim between the pilot study and Study 1, the Woodcock–Johnson III was published. The format and administration of the two versions of the Writing Fluency subtest are the same. The internal consistency reliability of the newer version, as assessed by a Rasch analysis, was .86, with a standard error of measurement of 7.19 in W-scale units and 5.63 in standard score units. Interrater agreement ranged from .98 to .99.
Digit Symbol Copy
To evaluate participants’ graphomotor speed on a task not confounded with phonologically loaded retrieval processes, we group administered the Digit Symbol Copy task from the WAIS–III (Wechsler, 1997). The par- ticipants were given 90 s to copy rows of simple symbols into rows of blank boxes immediately below them. The total score was derived by the number of clearly identifiable symbols out of 133 written before the time limit. As reported in the test manual, the test–retest reliability coefficient is .90. Interrater agreement was .99.
Procedure
The only difference between this and the pilot study, other than the changes in measures, was that this study took place over two sessions rather than one. As stated previously, some participants in the pilot study complained that there were too many tasks for one session.
In the first session, participants were told that they were going to watch a 20-min videotaped lecture on the psychology of problem solving and to take notes on the lecture using the two pieces of paper provided in the materials packet. They were also informed that they would have time to study their notes after viewing the lecture and told to make their notes as complete as possible. After the lecture was completed, participants were given 10 min to study their notes in preparation for the test. Once they finished studying, they were asked to complete the letter fluency, Digit Symbol Copy, and verbal fluency measures, in that order. The last task of the first session was the test. Participants were told they had 10 min to write “an organized summary about the psychology of problem solving.” In the second session, which took place 2 days after the first, participants com- pleted the VWM and compositional fluency tasks. The entire study took approximately 90 min.
Results
Prior to the pilot study, we had hypothesized that notes’ quality and VWM would be related to test performance and that transcription fluency, spelling, notes, and the identification of main ideas would be related to quality of notes (see Figure 1).
The results of the pilot study suggest that notes’ quality might directly mediate the relationship between the other independent variables and test performance. We tested the revised model (Figure 2), using a path analysis (AMOS 5, in SPSS, Release 11.0.1).
See Table 6 for the means and standard deviations of the dependent and independent variables and Table 7 for their intercorrelations. There was a ceiling effect with the Digit Symbol Copy subtest, so it was not included in the analyses. Parameter estimates for the model were generated via maximum likelihood estimation. Several indexes of fit are reported. The assumption of underlying bivariate normality was tested by the root-mean-square error of approximation (RMSEA) fit index. An RMSEA value lower than .05 indicates a close fit of the model relative to the degrees of freedom and no serious effects of nonnormality. The proportion of improvement in the fit of the model over the null model was evaluated with the normed fit index (NFI), the comparative fit index (CFI), and the Tucker–Lewis index (TLI), which is sometimes referred to as the nonnormed fit index. All three are interpreted in approxi- mately the same way, although the CFI is less affected by sample size than the NFI, and the TLI includes a correction for model complexity (and is the only one of those listed that can fall outside the range of .00 –1.00). All three should be greater than .95, which indicates that the fit of the researcher’s model is 95% better than the null model (in which the observed variables are assumed to be uncorrelated).
The goodness of fit indexes were very good, �2(5, N � 151) � 1.51, p � .91 (CFI � 1.00, RMSEA � .000, NFI � .993, TLI � 1.11). Path significance was based on critical ratios (CRs). A CR greater than 1.96 is considered to be significant at p � .05. The analysis indicated that notes’ quality predicted test performance (CR � 7.115, p � .001) and letter fluency predicted notes’ quality (CR � 2.96, p � .003). None of the other variables was signifi- cant. The overall model is presented in Figure 3. The CRs and other statistics are presented in Table 8.
Discussion
The results of Study 1 replicate the findings of the pilot study. Quality of notes was the only significant predictor of test perfor- mance, and transcription fluency, as measured by letter fluency, was the only significant predictor of quality of notes. Neither verbal fluency nor VWM contributed a significant amount of variance above that contributed by transcription fluency (as mea- sured by letter fluency). However, there was a ceiling effect with
Table 6 Study 1: Means and Standard Deviations
Statistic Recall Letter flu Notes Comp. flu VWM Sem. flu Phon. flu
M 7.45 57.32 20.94 27.51 4.70 33.00 28.10 SD 3.32 10.33 5.50 3.49 0.88 5.91 5.72
Note. flu � fluency; Comp. � composition; VWM � verbal working memory; Sem. � semantic; Phon. � phonetic.
175SKILL IN LECTURE NOTE TAKING
the Symbol Coding subtest of the WAIS–III4; thus, there was not enough variance to test whether a task that measures fine motor speed for symbols that is not verbally loaded would significantly predict quality of notes.
General Discussion
The primary purpose of the studies reported in this article is to evaluate the hypothesis that transcription fluency, VWM capacity, and the ability to identify main ideas would be related to the quality of notes. We found that transcription fluency (especially letter fluency) was a consistent predictor of notes’ quality. VWM was correlated with notes in the pilot study but not in Study 1 and was not found to be a unique contributor to notes’ quality. The ability to identify main ideas was not correlated with anything. Thus, the results of both studies extend the findings of the chil- dren’s and adult’s writing literature to suggest that transcription fluency is important not only to writing essays but to recording the ideas presented in lecture as well.
Although letter fluency was found to be a good predictor of notes’ quality, this finding is difficult to interpret, as discussed previously. First, research indicates that at least two skills contrib- ute variance to letter fluency among young children: fine motor speed and orthographic coding. Previous research suggests the surprising finding that fluency is more strongly correlated with the latter than with the former (Abbott & Berninger, 1993; Berninger et al., 2006; Berninger & Hooper, in press; Berninger & Richards, 2002). As discussed previously, we tried to measure both; the Wide Range Achievement Test Spelling subtest was used to mea- sure orthographic coding in the pilot study, and the Digit Symbol Copy subtest of the WAIS–III was used to measure fine motor speed in Study 1. The Spelling subtest was not a significant predictor, and problems with the Digit Symbol Copy subtest pre- vented us from measuring fine motor speed. Future research on the cognitive processes related to note taking should attempt to eval- uate the contribution of these processes using other measures.
There may be two reasons why our prediction that VWM would be related to notes’ quality was not upheld. First, VWM was significantly correlated with letter fluency in both studies, although more strongly in the first (r � .44) than in the second (r � .19), and letter fluency was more strongly correlated with notes in both studies than was VWM. In other words, VWM might not have accounted for a sufficient amount of unique variance to predict notes. This may indicate that letter fluency and VWM have a common underlying construct—the speeded access of verbal codes from long-term memory. This would include the phonological codes associated with letters in the letter fluency task and words in the VWM task. If so, this may indicate that differences among learners in VWM are due not to structural differences in capacity but to the quantity and quality of resources needed to process
4 We chose the Symbol Coding subtest because the task measured what we wanted it to measure and because it is part of a very well-standardized measure of intelligence. Thus, we assumed that the subtest would have the appropriate psychometric properties (e.g., normal distribution of scores). Unfortunately, it does not. According to the manual, 50% of the normative group obtained a score of 130 out of a possible 133. In our sample, 73.6% scored between 130 and 133, and 46.4% had a perfect score (133). Thus, for this population, this test produced results that were highly negatively skewed. If this test is used in the future with a sample comparable to the one used in this study, test time should be reduced significantly, and raw scores should be used in the data analysis.
letter fluency
seman ret
VWM
phon ret
.10
notes quality
comp fluency
error1
.26
.44
.37
.50
.30 .51
.11
.20
.44
.26
test perform
error2 .09
-.14
-.06
.29
.19
.51.09
Figure 3. Results of the evaluation of the relationship of letter and composition (comp) fluency, verbal working memory (VWM), phonetic retrieval (phon ret), and semantic retrieval (seman ret) to notes and the relationship of notes to test performance (test perform).
Table 7 Study 1: Intercorrelations Among the Independent and Dependent Variables
Variables 1 2 3 4 5 6 7 8
1. Recall — 2. Letter flu .12 — 3. Comp. flu .06 .50*** — 4. Notes .51*** .28*** .20* — 5. VWM �.10 .19* .20* �.02 — 6. Digit symbol .02 .31*** .11 .06 .19* — 7. Sem. flu .03 .51*** .44*** .08 .26*** .26*** — 8. Phonetic flu .08 .30*** .37*** .15 .11 .15 .44*** —
Note. flu � fluency; Comp. � composition; VWM � verbal working memory; Sem. � semantic. * p � .05. *** p � .01
176 PEVERLY ET AL.
information in VWM. Kintsch (1998), Perfetti (1986), and Vellu- tino (2001), for example, believed that reading comprehension skill is related to the quantity and quality of verbal resources pertaining to the interpretation of words, once word recognition is automatized, and not to capacity-related differences in VWM. Our finding also may mean, however, that we did not measure the right component of working memory. There is not a great deal of unanimity among researchers about what working memory is. Indeed, when referring to differences among theories of working memory, Kintsch, Healy, Hegarty, Pennington, and Salthouse (1999) stated that “it is rather difficult to identify a common core in terms of the phenomena under consideration” (p. 436). It should not come as a surprise, then, that there are at least four different categories of working memory theories (Miyake, 2001; Miyake & Shah, 1999; Peverly, 2006). In these theories, individual differ- ences are hypothesized to be due to structural differences in capacity (Just & Carpenter, 1992), the ability to attend (Engle, 2001, 2002), variation in the long-term memory resources needed to process information in VWM (Cowan, 1999; Ericsson & Kintsch, 1995), or all of the above (Baddeley, 2001). Although the kind of task used in this study is very similar to those used in other studies to measure variations in capacity and resources, it might not have been sensitive to variations in attention (Engle, 2001). Thus, future research should include two types of working memory tasks—the type used in this study, and the type used in Engle’s research on working memory as attention.
The second purpose of this study, to demonstrate that quality of notes was significantly and positively related to test performance, was upheld in both studies, which supports the findings of previous research (Bretzing & Kulhavy, 1981; Fisher & Harris, 1973; Kiewra, 1985; Kiewra et al., 1991; Kiewra & Fletcher, 1984; Peverly et al., 2003; Rickards & Friedman, 1978; Titsworth & Kiewra, 2004). Collectively, these data indicate that students’ representations of the structure and content of a lecture, as incom- plete as they often are (typically less than 40% of the information presented; e.g., Kiewra, DuBois, Christensen, Kim, & Lindberg, 1989), predict test performance better than variables that typically correlate quite well with overall school performance, such as verbal ability (Kiewra & Benton, 1988; Kiewra et al., 1987; Peverly et al. 1998) and GPA (Kiewra & Benton, 1988; Kiewra et al., 1987). In fact, research has found very few variables that predict test outcomes when notes (quantity and/or quality) are included among the predictor variables. The exceptions are back-
ground knowledge (Peper & Mayer, 1986; Peverly et al., 2003) and metacognitive judgments of learning, students’ judgments of how prepared they are to take a test or how well they did once they finished it (Peverly et al., 2003). What is also surprising is that none of the aforementioned variables has been found to correlate significantly with notes (Kiewra & Benton, 1988; Kiewra et al., 1987; Peverly et al. 1998, 2003). The exception, discussed previ- ously, is information processing ability (which some have labeled as VWM; Kiewra & Benton, 1988; Kiewra et al., 1987). Thus, proximal variables, those related to the processing of information pertaining to test content (lecture notes, background knowledge, and metacognitive judgments of how prepared students are to take a test), seem to be related more to test performance than are the distal variables that predict overall performance in school (prior GPA, SAT or Graduate Record Examination scores).
Conclusions and Implications
Contemporary views of cognitive processing and expertise (e.g., Anderson, 1990; Baddeley, 2000; Ericsson & Kintsch, 1995; Kintsch, 1998; Schneider & Shiffrin, 1977; Shiffrin & Schneider, 1977) argue that learning skills, including many school-based tasks, such as reading and writing, depend on performing a hier- archy of skills simultaneously (in parallel). In the execution of these skills, at least three conditions must hold. First, domain- specific basic skills must be executed with an acceptable degree of fluency or automaticity, so that most, if not all, of the space in working memory can be used for the application of the higher level cognitive skills needed to produce successful outcomes. If basic skills are not automatized, the application of higher level cognitive skills can be attenuated and prevent students from achieving their goal, even if their cognitive and metacognitive resources are sub- stantial. Second, as implied in the previous sentence, individuals must have the cognitive resources (knowledge, strategies, execu- tive monitoring) necessary to enable them to attend, interpret, and process the information in VWM once basic skills become autom- atized. Finally, individuals must have the VWM capacity neces- sary to process information adequately.
Data from these studies suggest that the basic skill of transcrip- tion fluency is related to quality of notes. Faster transcription fluency enables students to record more and higher quality infor- mation from a lecture. These data also suggest that VWM is not independently related to skill in note taking. However, this should be verified in future research with different complex span tasks, given the lack of consensus among researchers about what such tasks actually measure (Daneman, 2001). Also, future research should measure note takers’ selective attention, as Engle (2002) argued that capacity is related to the “ability to control attention [and avoid distraction] to maintain information in an active, quickly retrievable state” (p. 20). It may be the ability to attend, not the capacity of VWM, that partially accounts for skill in taking notes. Finally, the main idea task (pilot study) did not contribute to the skill of note taking. Logically, some variable must be related to the ability to identify and construct important information during a lecture. Future researchers may want to use a listening rather than a reading comprehension task. Although both measure the same higher level cognitive processes (Kintsch, 1998), the former is not confounded by differences in word recognition speed.
Table 8 Study 1: Summary of the Structural Equation Model
Structural path Estimate SE CR
Comp. flu 3 notes .09 .15 0.95 Sem. flu 3 notes �.14 .09 �1.36 VWM 3 notes �.06 .51 �0.76 Let. flu 3 notes .29 .05 2.96**
Phon. flu 3 notes .09 .09 1.04 Notes 3 essay .51 .04 7.12****
Note. CR � critical ratio; Comp. � composition; flu � fluency; Sem. � semantic; VWM � verbal working memory; Let. � letter; Phon. � phonetic. ** p � .01. **** p � .001.
177SKILL IN LECTURE NOTE TAKING
The findings from the pilot study and Study 1 on the relationship of transcription fluency to notes’ quality may have important educational implications. First, systematic instruction in handwrit- ing in elementary school might have a positive effect not only on the quantity and quality of essays written by children in elementary and middle school (Berninger et al., 1997; Graham et al., 2000; Jones & Christensen, 1999) but on the quality of notes taken by high school and college students. Longitudinal research is needed to evaluate this conjecture. Second, a transcription fluency com- ponent (among other components) should be included in instruc- tion on lecture note taking to evaluate whether it can improve older (high school and college) students’ handwriting fluency and whether improvements in fluency result in higher quality notes.
References
Abbott, R., & Berninger, V. (1993). Structural equation modeling of relationships among developmental skills and writing skills in primary and intermediate grade writers. Journal of Educational Psychology, 85, 478 –508.
Anderson, J. R. (1982). Acquisition of cognitive skill. Psychological Re- view, 89, 369 – 406.
Anderson, J. R. (1990). The adaptive character of thought. Hillsdale, NJ: Erlbaum.
Baddeley, A. D. (1998). Human memory: Theory and practice. Boston: Allyn & Bacon.
Baddeley, A. D. (2000). The episodic buffer: A new component of working memory? Trends in Cognitive Sciences, 4, 417– 423.
Baddeley, A. D. (2001). Is working memory still working? American Psychologist, 56, 849 – 864.
Bayliss, D. M., Jarrold, C., Gunn, D. M., & Baddeley, A. D. (2003). The complexities of complex span: Explaining individual differences in working memory in children and adults. Journal of Experimental Psy- chology: General, 132, 71–92.
Berninger, V. W. Abbott, R. D., Jones, J., Wolf, B. J., Gould, L., Anderson- Youngstrom, M., et al. (2006). Early development of language by hand: Composing-, reading-, listening-, and speaking-connections, three letter writing modes, and fast mapping in spelling. Developmental Neuropsy- chology, 29, 61–92.
Berninger, V. W. & Hooper (in press). A developmental perspective on writing disabilities in children and youth. In D. Molfese & V. Molfese (Eds.), Handbook of child neuropsychology. Mahwah, NJ: Erlbaum.
Berninger, V. W, Mizokawa, D., & Bragg, R. (1991). Theory based diagnosis of remediation of writing disabilities. Journal of School Psy- chology, 29, 57–79.
Berninger, V. W., & Richards, T. L. (2002). Brain literacy for educators and psychologists. San Diego, CA: Academic Press.
Berninger, V. W., Vaughan, K. B., Abbott, R. D., Abbott, S. P., Rogan, L. W., Brooks, A., et al. (1997). Treatment of handwriting problems in beginning writers: Transfer from handwriting to composition. Journal of Educational Psychology, 89, 652– 666.
Berninger, V. W., Vaughan, K. B., Abbott, R. D., Brooks, A., Abbott, S. P., Rogan, L. W., et al. (1998). Early intervention for spelling problems: Teaching functional spelling units of varying size with a multiple- connections framework. Journal of Educational Psychology, 90, 587– 605.
Bretzing, B. H., Kulhavy, R. W. (1981). Notetaking and passage style. Journal of Educational Psychology, 73, 242–250.
Brobst, K. E. (1996). The process of integrating information from two sources, lecture and text. (Doctoral dissertation, Teachers College, Co- lumbia University, 1996). Dissertation Abstracts International, 57, 217.
Brown, A. L., & Day, J. D. (1983). Macrorules for summarizing text: The development of expertise. Journal of Verbal Learning and Verbal Be- havior, 22, 1–14.
Brown, A. L., Day, J. D., & Jones, R. S. (1983). The development of plans for summarizing texts. Child Development, 54, 968 –979.
Brown, J. S., McDonald, J. L., Brown, T. L., & Carr, T. H. (1988). Adapting to processing demands in discourse production: The case of handwriting. Journal of Experimental Psychology: Human Perception and Performance, 14, 45–59.
Chi, M. T. H., Glaser, R., & Farr, M. (Eds.). (1988). The nature of expertise. Hillsdale, NJ: Erlbaum.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Cohn, E., Cohn, S., & Bradley, J. (1995). Notetaking, working memory, and learning in principles of economics. Research in Economic Educa- tion, 26, 291–307.
Connelly, V., Campbell, S., MacLean, M., & Barnes, J. (2006). Contribution of lower order skills to the written composition of college students with and without dyslexia. Developmental Neuropsychology, 29, 175–196.
Connelly, V., Dockrell, J. E., & Barnett, J. (2005). The slow handwriting of undergraduate students constrains overall performance in exam es- says. Educational Psychology, 25, 97–105.
Cowan, N. (1999). The embedded-process model of working memory. In A. Miyake & P. Shah (Eds.), Models of working memory: Mechanisms of active maintenance and executive control (pp. 62–101). New York: Cambridge University Press.
Daneman, M. (2001). Learning disabled individuals show deficits on working memory tasks: The question is why? Issues in Education, 7, 79 – 85.
Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior, 19, 450 – 466.
Daneman, M., & Carpenter, P. A. (1983). Individual differences in inte- grating information between and within sentences. Journal of Experi- mental Psychology: Learning, Memory, and Cognition, 9, 561–583.
Daneman, M., & Hannon, B. (2001). Using working memory theory to investigate the construct validity of multiple-choice reading comprehen- sion tests such as the SAT. Journal of Experimental Psychology: Gen- eral, 130, 208 –223.
Di Vesta, F. J., & Gray, S. G. (1973). Listening and note taking: II. Immediate and delayed recall as functions of variations in thematic continuity, note taking, and length of listening-review intervals. Journal of Educational Psychology, 64, 278 –287.
Dunkel, P., & Davy, S. (1989). The heuristic of lecture notetaking: Per- ceptions of American and international students regarding the value and practice of notetaking. English for Specific Purposes, 33–50.
Dunkel, P., Mishra, S., & Berliner, D. (1989). Effects of note taking, memory, and language proficiency on lecture learning for native and non-native speakers of English. TESOL Quarterly, 23, 543–549.
Engle, R. W. (2001). What is working memory capacity? In H. L. Roedi- ger, J. S. Nairne, I. Neath, & A. M. Suprenant (Eds.), The nature of remembering: Essays in honor of Robert G. Crowder (pp. 297–314). Washington, DC: American Psychological Association.
Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological Science, 11, 19 –23.
Ericsson, K. A., & Kintsch, W. (1995). Long-term working memory. Psychological Review, 102, 211–245.
Fisher, J. L., & Harris, M. B. (1973). Effect of note-taking and review on recall. Journal of Educational Psychology, 65, 321–325.
Graham, S., Berninger, V. W., Abbott, R. D., Abbott, S. P., & Whitaker, D. (1997). Role of mechanics in composing of elementary school students: A new methodological approach. Journal of Educational Psychology, 89, 170 –182.
Graham, S., Harris, K. R., & Fink, B. (2000). Is handwriting causally related to learning to write? Treatment of handwriting problems in beginning writers. Journal of Educational Psychology, 92, 620 – 633.
Jones, D., & Christensen, C. A. (1999). Relationship between automaticity
178 PEVERLY ET AL.
in handwriting and students’ ability to generate text. Journal of Educa- tional Psychology, 91, 44 – 49.
Just, M. A., & Carpenter, P. A. (1992). A capacity theory of comprehen- sion: Individual differences in working memory. Psychological Review, 99, 122–149.
Kellogg, R. T. (2001). Competition for working memory among writing processes. American Journal of Psychology, 114, 175–191.
Kellogg, R. T. (2004). Working memory components in written sentence generation. American Journal of Psychology, 117, 341–361.
Kiewra, K. A. (1985). Investigating notetaking and review: A depth of processing alternative. Educational Psychologist, 20, 23–32.
Kiewra, K. A., & Benton, S. L. (1988). The relationship between infor- mation processing ability and notetaking. Contemporary Educational Psychology, 13, 33– 44.
Kiewra, K. A., Benton, S. L., & Lewis, L. B. (1987). Qualitative aspects of notetaking and their relationship with information-processing ability and academic achievement. Journal of Instructional Psychology, 14, 110 –117.
Kiewra, K. A., DuBois, N. F., Christensen, M., Kim, S.-I., & Lindberg, N. (1989). A more equitable account of the note-taking functions in learn- ing from lecture and from text. Instructional Sciences, 18, 217–232.
Kiewra, K. A., DuBois, N. F., Christian, D., McShane, A., Meyerhoffer, M., & Roskelley, D. (1991). Note-taking functions and techniques. Journal of Educational Psychology, 83, 240 –245.
Kiewra, K. A., & Fletcher, H. J. (1984). The relationship between levels of note-taking and achievement. Human Learning: Journal of Practical Research & Applications, 3, 273–280.
Kintsch, W. (1998). Comprehension: A paradigm for cognition. Cam- bridge, England: Cambridge University Press.
Kintsch, W., Healy, A. F., Hegarty, M., Pennington, B. F., & Salthouse, T. A. (1999). Models of working memory: Eight questions and some general issues. In A. Miyake & P. Shah (Eds.), Models of working memory (pp. 412– 441). Cambridge, England: Cambridge University Press.
Kline, R. (1998). Principles and practice of structural equation modeling. New York: Guilford Press.
Kobayashi, K. (2005). What limits the encoding effect of note-taking? A meta-analytic examination. Contemporary Educational Psychology, 30, 242–262.
Korkman, M., Kirk, U., & Kemp, S. (1998). NEPSY: A developmental neuropsychological assessment. San Antonio, TX: Psychological Cor- poration.
McCutchen, D., Covill, A., Hoyne, S. H., & Mildes, K. (1994). Individual differences in writing: Implications of translating fluency. Journal of Educational Psychology, 86, 256 –266.
McIntyre, S. (1992). Lecture notetaking, information-processing, and aca- demic achievement. Journal of College Reading and Learning, 25, 7–17.
Meyer, B. J. F. (1985). Prose analysis: Purposes, procedures, and problems. In B. K. Britton & J. B. Black (Eds.), Understanding expository text: A theoretical and practical handbook for analyzing expository text (pp. 11– 64). Hillsdale, NJ: Erlbaum.
Meyer, B. J. F., & Poon, L. W. (2000). Effects of structure strategy training and signaling on recall of text. Journal of Educational Psychology, 93, 141–159.
Miyake, A. (2001). Individual differences in working memory: Introduc- tion to the special issue. Journal of Experimental Psychology: General, 130, 163–168.
Miyake, A., & Shah, P. (Eds.). (1999). Models of working memory. Cambridge, England: Cambridge University Press.
Olive, T., & Kellogg, R. T. (2002). Concurrent activation of high- and low-level production processes in written composition. Memory & Cog- nition, 30, 594 – 600.
Palmatier, R. A., & Bennett, J. M. (1974). Notetaking habits of college students. Journal of Reading, 18, 215–218.
Peper, R. J., & Mayer, R. E. (1986). Generative effects of note-taking during science lectures. Journal of Educational Psychology, 78, 34 –38.
Perfetti, C. A. (1986). Cognitive and linguistic components of reading ability. In B. Foorman & A. W. Siegel (Eds.), Acquisition of reading skills: Cultural constraints and cognitive universals (pp. 11– 40). Hills- dale, NJ: Erlbaum.
Peters, D. L. (1972). Effects of note taking and rate of presentation on short-term objective test performance. Journal of Educational Psychol- ogy, 63, 276 –280.
Peterson, L. R., & Peterson, M. J. (1959). Short-term retention of individ- ual verbal items. Journal of Experimental Psychology, 58, 193–198.
Peverly, S. T. (2006). The importance of handwriting speed in adult writing. Developmental Neuropsychology, 29, 197–216.
Peverly, S. T., Brobst, K., Graham, M., & Shaw, R. (2003). College adults are not good at self-regulation: A study on the relationship of self- regulation, note-taking, and test-taking. Journal of Educational Psychol- ogy, 95, 335–346.
Peverly, S., Brobst, K. E., Shaw, R., & Graham, M. (1998, May). Rela- tionship of verbal ability, note-taking, test type, and monitoring to test performance. Poster session presented at the 10th Annual Meeting of the American Psychological Society, Washington, DC.
Piolat, A., Olive, T., & Kellogg, R. T. (2005). Cognitive effort during note taking. Applied Cognitive Psychology, 19, 291–312.
Putnam, M. L., Deshler, D. D., & Schumaker, J. B. (1993). The investigation of setting demands: A missing link in learning strategy instruction. In L. S. Meltzer (Ed.), Strategy assessment and instruc- tion for students with learning disabilities (pp. 325–354). Austin, TX: PRO-ED.
Ransdell, S., & Levy, C. M. (1996). Working memory constraints on writing quality and fluency. In C. M. Levy & S. Ransdell (Eds.), The science of writing (pp. 93–105). Mahwah, NJ: Erlbaum.
Ransdell, S., Levy, C. M., & Kellogg, R. T. (2002). The structure of writing processes as revealed by secondary task demands. Educational Studies in Language and Literature, 2, 141–163.
Rayner, K., Foorman, B. R., Perfetti, C. A., Pesetsky, D., & Seidenberg, M. S. (2001). How psychological science informs the teaching of read- ing. Psychological Science in the Public Interest, 2, 31–74.
Rickards, J. P., & Friedman, F. (1978). The encoding versus the external storage hypothesis in note taking. Contemporary Educational Psychol- ogy, 3, 136 –143.
Sattler, J. M. (2001). Assessment of children: Cognitive applications. La Mesa, CA: Jerome M. Sattler.
Schneider, W., & Shiffrin, R. M. (1977). Controlled and automatic human information processing: I. Detection, search, and attention. Psychologi- cal Review, 84, 1– 66.
Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human information processing: II. Perceptual learning, automatic attending, and a general theory. Psychological Review, 84, 127–190.
Swanson, H. L., & Berninger, V. (1996). Individual differences in chil- dren’s working memory and writing skills. Journal of Experimental Psychology, 63, 358 –385.
Swanson, H. L., & Siegel, L. (2001). Learning disabilities as a working memory deficit. Issues in Education, 7, 1– 48.
Thomas, J. W., Iventosch, L., & Rohwer, W. D. (1987). Relationships among student characteristics, study activities, and achievement as a function of course characteristics. Contemporary Educational Psychol- ogy, 12, 344 –364.
Titsworth, B. S., & Kiewra, K., A. (2004). Spoken organizational lecture cues and student notetaking as facilitators of student learning. Contem- porary Educational Psychology, 29, 447– 461.
Turner, M. L., & Engle, R. W. (1989). Is working memory capacity task dependent? Journal of Memory and Language, 28, 127–154.
Vellutino, F. R. (2001). Working memory deficits and learning disabilities: Reactions to Swanson and Siegel. Issues in Education, 7, 49 – 69.
179SKILL IN LECTURE NOTE TAKING
Vellutino, F. R., Fletcher, J. M., Snowling, M. J., & Scanlon, D. M. (2004). Specific reading disability (dyslexia): What have we learned in the past four decades? Journal of Child Psychology and Psychiatry, 45, 2– 40.
Voss, J. (1989). Problem solving and the educational process. In A. Lesgold & R. Glaser (Eds.), Foundations for a psychology of education (pp 251–294). Hillsdale, NJ: Erlbaum.
Wade, S. E., & Trathen, W. (1989). Effect of self-selected study methods on learning. Journal of Educational Psychology, 81, 40 – 47.
Wechsler, D. (1997). Wechsler Adult Intelligence Scale—Third Edition. San Antonio, TX: Psychological Corporation.
Wilkinson, G. S. (1993). Wide Range Achievement Test—Revision 3. Wilmington, DE: Jastak Associates.
Woodcock, R. W., & Johnson, M. B. (1989). Woodcock–Johnson Psycho- educational Battery—Revised. Itasca, IL: Riverside.
Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). The Woodcock– Johnson III. Itasca, IL: Riverside.
Appendix
The Structure and Content of the Essay
I. Functions of problem solving in education a. Problem solving is a cognitive activity important to
educational theory and classroom practice.a
b. Problem solving is considered part of learning subject matter. It serves a testing and teaching function.a
II. Definition of a problem a. A problem is said to exist when an individual has a
particular goal but is unable to obtain that goal. b. It is frequently assumed that there is some type of
obstacle or barrier that prevents the solver from reaching the goal.
c. These obstacles must, of necessity, be broadly defined and include such factors as failure to remember and lack of information.a
III. Information processing approach a. Concepts
1. Problem representationa
2. Goal statesa
3. Constraintsa
4. Problem statesa
5. Operatorsa
6. Ill-structured problemsa
b. Example—Tower of Hanoia
IV. Research findings: Problem solving in particular domains a. Chessa
b. Physicsa
V. Factors involved in problem solving a. Understanding the problem representationa
b. Effective problem solving is related to abstract knowledge structuresa
VI. Instructability of general problem solvinga
a Indicates separate content areas.
Received July 15, 2005 Revision received June 29, 2006
Accepted July 6, 2006 �
180 PEVERLY ET AL.