Return to index.

Denise Elmer*

Wasted Time or Informed Delay?
Academic Procrastination, Learning Style, and Self-Talk


This research determined if Kolb’s (1976) learning style preferences could be used to identify the personality makeup of academic procrastinators. In addition, intrapersonal communication, or self-talk, as related to procrastination and learning style preference was identified. The 174 participants, 101 males and 65 females, were students enrolled in public speaking classes. The measures for this study were self-report procrastination and learning style surveys published by Tuckman (1991) and Kolb (1976). Results indicated that procrastination between learning style preferences is distinctly different for males. Academic procrastination can be predicted for students preferring active and concrete learning. It is evident from this study that content analysis of significant survey questions can be interpreted as self-talk, providing a basis for qualitative exploration during focus group/interview sessions. Survey self-talk exhibits evidence of poor self-image and self-blame with an inability to integrate courses of action that avoid procrastination. Female students’ self-talk reveals task avoidance and negative self-esteem, whereas male students describe a tendency to avoid tasks, negative self-esteem, and self-blame.


Procrastination is a behavior in which we all indulge. Those of us with low procrastination tendencies find procrastinators annoying; however, because we all do it, this irritating behavior is often tolerated. Although most research on procrastination is in the area of psychology, procrastination does have a meaningful relationship to the field of communication; it is a strategy used after self-talk has occurred.

Self-talk is intrapersonal communication and can be described as a cognitive control process that involves the ability to think into the future, to place value on a choice, to like or dislike a choice, to comprehend the relationship(s) between cause and effect, and to control the choice made (Roberts, Edwards, & Barker, 1987). Self-talk centers around self-concept or the positive and negative views one has of one’s physical, psychological, and social self. Another term for this self-view is self-esteem. Knaus (1995), a pioneer in procrastination research, counsels clients to “listen for anxiety-creating self-talk” when starting to procrastinate and encourages clients to not only “listen for self-downing self-talk,” but to confront “negative self-talk” (par. 52-53). Thus, procrastination is an outcome of negative self-talk.

Academic procrastination is a form of procrastination peculiar to education. Research indicates that academic procrastination is associated with communication apprehension, ineffective learning strategies, lower grade point averages, boring or difficult assignments, unplanned study habits, fraudulent excuses, plagiarism, anxiety, fear of failure, depression, irrational thought, less self-efficacy, less self-control, and less delay of gratification (Behnke & Sawyer, 1999; Chissom & Iran-Nejad, 1992; Ferrari, Johnson, & McCown, 1995; Ferrari & Beck, 1999; Haycock, McCarthy, & Skay, 1998; Lay, 1986; Milgram, 1995; Roig & DeTommaso, 1995; Senecal, 1995; Solomon & Rothblum, 1984). Ferrari, Parker, and Ware (1992) concluded that behavioral outcomes due to procrastination “encompass both cognitive and affective personality components requiring further analysis” (p. 496). Their suggested solution is to identify the personality “makeup” of procrastinators.

Most procrastination investigations focus on personality attributes such as self-efficacy. However, a test for self-efficacy is less likely to be used in college classrooms because the results have little relation to the topic being learned. On the other hand, learning style surveys are being administered more often in classrooms. In the instance of this research, the use of learning styles is taught as a public speaking technique to students, while instructors are encouraged to use survey results as a curriculum device (Sellnow, 1999a, 1999b). Kolb (1976) defined learning style as an individual’s preferred way to learn. Curry (1987) enhanced this description by describing learning style as a cognitive personality process that adapts and assimilates information. Assimilation centers on the use of learning strategies or “translation-like mechanisms” that individuals use to “cope with” the environment (p. 5). Chissom and Iran-Nejad (1992) determined that procrastination is a strategy that students use to delay learning. Thus, the literature indicates that procrastination is both a behavior and a strategy resulting from negative self-talk.

Procrastination and self-efficacy
The earliest use of procrastination had one definition, an “informed delay,” and was thought to be an intelligent choice (Ferrari et al., 1995, p. 4). With the onset of the Industrial Revolution a more negative connotation, “sloth,” became popular (p. 4). Sloth denotes avoiding work, and includes manipulating others to do it for you. Ferrari et al. wrote that with modern-day economic growth, the term “sloth” changes to “task avoidance,” and described procrastination as irrational thought which originates in cognitive theory (p. 5). Characteristics of irrational thought include illogical thoughts, self-defeating choices that undermine goals, fear of failure, immediate gratification, and too high standards or perfectionism.

Haycock et al. (1998) tested if procrastination can be predicted by efficacy expectations using a modified Form G Procrastination Inventory. Haycock et al. operationalized efficacy using Bandura’s (1986) theory—[efficacy is a] “generative capability in which cognitive, social, and behavior subskills must be organized into integrated courses of action” (p. 391). Bandura noted that “self-doubters” abandon this generative process sooner than “persevering” individuals (p. 391). Lay (1986), developer of Form G, confirmed that irrational thought is related to procrastination by concluding that some people judge time requirements inaccurately, resulting in procrastination. Describing this thought process as irrational; Lay related a lack of time judgment to disorganization and an inability to maintain priorities. This conclusion is similar to Bandura’s definition of self-efficacy as well.

To measure efficacy, Haycock et al. (1998) created the Self-Efficacy Inventory (SEI) which assesses “behaviors related to the task of doing an important and difficult project by a specific deadline” (p. 319). Through regression analysis, Haycock et al. found students with low self-efficacy self-report more procrastination on all the behaviors needed to accomplish a major project. Haycock et al. claimed that their findings support Bandura’s argument that “strong efficacy expectations lead to greater task initiation and persistence, whereas weak expectations produce task avoidance and less persistence” (p. 321).

According to Tuckman (1989) each of us has a tendency to procrastinate. He suggested that this proclivity is due to a preference for “ease and pleasure over hard work and pain,” a conviction that trying will lead to failure or “self-doubt,” and a capacity to blame external people and events for our neglect (pp. 48-49). Tuckman’s theory originates in Bandura’s (1986) social cognitive theory. This theory assumes that human behavior is due to the interaction of personal behavior, personal cognition, and external events. This interaction occurs because of people’s ability to translate external symbols into knowledge that regulates future behavior. However, as humans make mistakes, this ability does not preclude flawed assessments from being formed and acted on. Behavior is both motivated and self-regulated by comparing the desired behavior outcome to internal standards and regulations. Based on this construct of self-efficacy, academic procrastination may be the result of self-talk that includes a self-capability judgment to do a particular educational task.

Tuckman (1991) devised a self-report measure of academic procrastination tendencies for the purpose of providing college students a self-report scale that may help “regulate their own learning” and obtain a “behavioral measure of self-regulated performance with self-efficacy or the degree to which one believes in one’s own capability to perform” (p. 474). This survey is a 35-item (TPS35), 4-point, Likert-type scale with a reliability coefficient of .90. Further testing reduced it to a 16-item scale (TPS16), with a reliability coefficient of .86. The Tuckman Procrastination Scale (TPS) has a higher reliability score than either the two-part PASS, .26, .81; or Aitken Procrastination Inventory (API), .82, two other common measures of student procrastination (Ferrari et al., 1995).

The TPS is designed to measure (1) a generalized self-description of the inclination to delay, (2) a penchant to experience difficulty when doing “unpleasant things,” and (3) a tendency to blame external events or people for any ensuing ramifications from procrastination choices (Tuckman, 1991, p. 475). Survey questions Tuckman used to illustrate these dimensions are, respectively, (1) “When I have a deadline, I wait until the last minute,” (2) “I look for a loophole or a shortcut to get through a tough task,” and (3) “I believe that other people do not have the right to give me deadlines” (p. 475).

To measure self-doubt or self-efficacy, Tuckman (1991) correlated TPS responses to the General Self-Efficacy Scale (GSE). The purpose of the GSE is to measure people’s belief in “their ability to overcome obstacles and succeed” (p. 476). Examples of GSE questions are
(1) “Failure just makes me try harder,” and (2) “I feel insecure about my ability to do things”
(p. 476). Tuckman concluded that the TPS is a highly reliable measure of academic procrastination as related to self-efficacy. He described characteristics of procrastinators as

(1) “time wasters, deadline avoiders, pleasure seekers, and blamers and resenters of others who make demands on their time,” and as (2) self-doubters who worry about their capability to perform and rarely expend extra effort on tasks that would provide additional benefits such as graded bonus points (p. 479).

Learning style theory and self efficacy
Bandura’s (1986) social cognitive theory is the “conception-matching process whereby symbolic representations are translated into appropriate courses of actions” (p. 390). However, it does not explain how this practical learning is acquired. Examination of Kolb’s (1976) experiential learning theory indicates that learning is a four-stage cycle that moves from observations and reflections, to formation of abstract concepts and generalizations, to testing implication of concepts in new situations, to concrete experience, which again leads to observations and reflections—or as Kolb summarized, “the central process of human adaptation to the social and physical environment” (p. 6). These four phases can be explained by what Kolb described as the “core of the model,” which is how “experience is translated into concepts which in turn are used as guides in the choice of new experiences” (p. 2). In other words, the learning cycle is a marriage of learning and problem solving (Kolb, Rubin, & McIntyre, 1971).

Note the similarity of Bandura and Kolb’s theories. According to Bandura, self-efficacy is dependent on the ability to recognize external experiences, whereas, Kolb decided learning begins with the ability to recognize external symbols. Both abilities integrate these external symbols into concepts based on past values and experiences, thus producing guides. These guides self-regulate behavior and suggest new experiences, resulting in action or, as in the case of procrastination, avoiding action.

Kolb (1976) based his theory on the Jungian concept that “fulfillment in adult development is accomplished by higher level integration and expression of nondominant modes of dealing with the world” (p. 2). This learning model consists of two dimensions with polar opposites: (1) concrete experience/abstract conceptualization (CE/AC) and (2) active experimentation/reflective observation (AE/RO). The CE/AC dimension stimulates cognitive growth and learning, including the ability “to plan ahead ideationally and think and perform symbolically” (p. 3). The AE/RO dimension represents the ability to apply learning actively after having tested (reflected) it against internal standards. On the CE/AC dimension are two distinct learning types, Divergers (CE/feel) and Convergers (AC/think). The AE/RO dimension represents the two distinct learning types of Accommodators (AE/do) and Assimilators (RO/watch). To determine learning style, Kolb devised a Learning Style Inventory (LSI) that ranks an individual’s preferred way to learn. Hence, student learning can begin at any point within the four-stage cycle; it depends on the individual’s learning style personality characteristics.

Each learning style personality has assigned distinctive descriptors. Divergers’ (CE/feel) strength is action after reflection; they are imaginative and emotional (Kolb, 1976). Convergers’ (AC/think) strength is practical application of ideas; they are unemotional and focused on single solutions. Accommodators’ (AE/do) strength is to do. They are intuitive, preferring to adapt to the immediate situation and using trial and error to solve problems. Assimilators (RO/watch) prefer the creation of theoretical concepts with a lack of concern for the actual application of theory.

Procrastination, learning style, self-efficacy and self-talk
Perhaps the relationship between procrastination, learning style, self-efficacy and self-talk has its best explanation within the typology established by Curry’s (1987) research, which examined the reliability and validity of major North American, Australian, and European learning style instruments. Using psychometric evidence, Curry reorganized twenty-one learning style instruments into a system with three layers. Curry likened this system to an onion with the inner layer, or core, representing the “central personality dimension” (Hickcox, 1995, p. 28). The middle layer or center focused on a “theme of information-processing,” and the outer layer is that which has immediate interaction and is influenced by the environment (pp. 28-29). Curry associated the Myers-Briggs Type Indicator test (MBTI) with the inner core and the LSI with the middle layer. The middle layer, or information-processing dimension, represents the cognitive ability of an individual to “intersect personality levels, individual differences and environmentally based learning format choices” (p. 32). This characterization is similar to Bandura’s (1986) interpretation of self-efficacy, “the processes governing the interrelationship between knowledge and action” (p. 390), and Roberts et al. (1987) construct of self-talk, a cognitive control process that involves thinking into the future, placing value on a choice, liking or disliking a choice, comprehending the relationship(s) between cause and effect, and controlling the choice made.

Figure One illustrates how the information-processing dimension encompasses the interrelationship between self-efficacy, experiential learning, procrastination and self-talk. Students are exposed to external symbols in the way of textbooks and classroom board notes, and experiences through assignments and interaction with instructors and classmates. How this information is acquired depends on their individual learning style type or strength. Once external knowledge is acquired, it is processed internally by a cognitive comparison of observations, concepts, and values, resulting in self-regulated behavior and new experience choices. If self-efficacy is high, the academic objective is accomplished; if self-efficacy is low, the academic objective is avoided, i.e., procrastination occurs.

The purpose of this research was to test if Kolb’s (1976) learning style preferences could be used to identify the personality makeup of procrastinators. As there has been limited re-testing of the TPS, another purpose was to re-test the reliability of the scale (Ferrari et al., 1995). Lastly, this study examines how a self-completed survey is an articulation of self-talk.

A comparison of MBTI and self-regulated performance and procrastination characteristics to learning style characteristics suggests that a student preferring one learning style dimension is unlikely to procrastinate more than another, as both dimensions have procrastination characteristics (Briggs Myers, 1993; Ferrari et al. 1992; Kolb, 1976, 1981; Tuckman, 1991, 1998). However, prior personality makeup research did not consider negative self-efficacy as a factor (Ferrari et al. 1992). Tuckman (1991) defined procrastination as “the lack or absence of self-regulated performance; the tendency to avoid or put off an activity under one’s control” and established that procrastination is a behavior related to low self-efficacy (p. 474). Self-efficacy and learning style preference are interactions of self-talk or personal behavior, personal cognition, and external events; therefore, it is probable that learning style type can predict procrastination.

Within the AE/RO (do/watch) dimension, reflection inhibits action, and inversely, action inhibits reflection (Kolb, 1976); therefore, an Assimilator (RO/watch) learning style type is more likely to procrastinate than an Accommodator (AE/do). Within the CE/AC (feel/think) dimension, a Converger (AC/think) is more likely to “perform symbolically” and a Diverger (CE/feel) to experience (Kolb, 1976); therefore, a Converger learning style type is more likely to procrastinate. Alternatively, a combined preference for the two styles most inclined to procrastinate, Assimilator and Converger, may be more indicative of procrastination. Lastly, it appears that the Assimilator and Converger preferences may be prone to negative self-efficacy. Rank order then, from most likely to procrastinate, should be Assimilator due to preference for inhibition of action, Converger due to preference for symbolic performance, Diverger due to preference to experience, and Accommodator due to preference to carry out plans.
Therefore, because it is probable that learning style preference is related to procrastination, there are two hypotheses for this research.

Hypothesis One: students with the learning styles of Assimilator (RO/watch) and Converger (AC/think) are more likely to academically procrastinate than students with the styles of Accommodator (AE/do) and Diverger (CE/feel). Within the combined types, RO/AC (watch/think) is more likely to procrastinate CE/AE (feel/do).

Hypothesis Two: the rank order, from highest to lowest, of learning style academic procrastination tendency, is Assimilator, Converger, Diverger, and Accommodator.

Questions asked
Individual procrastination questions were tested to assess what variety of self-talk occurred within each learning style. It should be noted that these questions have not been interpreted as self-talk in other studies; however, the items measured in these questions fit the definition of intrapersonal communication (Roberts et al., 1987). There was no prediction as to which questions would be significant for each type.

The 174 participants in this study were students enrolled in public speaking classes at a Midwest, land-grant, state university. Classes are capped at twenty-two students each, with approximately 800 enrolled each semester. Public speaking is a required class for all students planning to graduate from this university of 9,000-plus students. Participants were invited to take part by their instructors and received extra credit points for completing a procrastination and learning style inventory. Confidentiality of students’ identities was maintained by use of control identification numbers. The LSI is a regular class assignment, usually completed outside of the classroom, although some instructors had students complete the surveys in class. In either instance, the purpose of the survey and instructions were covered orally with students.

Research instruments
The measures for this study were two survey instruments: the first a self-report procrastination survey (TPS) published by Tuckman (1991), and the second a self-report learning style survey (LSI) published and copyrighted by Kolb (1976). If a respondent left a reply blank or scores did not fall within defined ranges, the response, as well as the respondent’s corresponding TPS or LSI, was eliminated from the study. Ultimately, eight responses were rejected.

Procrastination instrument
The TPS is a 35-item, 4-point, Likert-type scale with a reliability of .90 (Tuckman, 1991). Using factor analysis, Tuckman reduced this 35-item scale to a 16-item scale consisting of questions that loaded .40 or higher with an alpha reliability of .86. The 16-item scale takes less time to complete than the 35-item scale and can be more quickly used by students for self-assessment of procrastination tendencies. The 16-point scale is contained within the 35-item scale, so both were administered at the same time. Self-efficacy and self-regulated behaviors are measured by default (Tuckman, 1991).

Methods used to re-test the reliability of the scale were alpha reliability and factor analysis. As in Tuckman’s study, factors loading .40 and higher were examined for reliability and interpretation of procrastination as related to self-efficacy and self-regulated behavior. A minimum threshold of .80 was predetermined to be an acceptable alpha reliability.

On the TPS instrument, students indicated their self-rated level procrastination tendency by selecting “1, That’s me,” “2, That’s my tendency,” “3, Not my tendency,” or “4, Not me.” The minimum score possible on the TPS 35-item scale is 35, and describes a high degree of self-reported procrastination. The maximum score possible is 140, and describes a non-procrastinator. Minimum and maximum scores for the TPS16 item scale are 16 and 64.

Each procrastination question with its corresponding numerical response was entered as a variable, as well as the total procrastination score, i.e., the sum of the 35 questions and the sum of the 16 questions. Factor analysis resulted in a new scale, PS14, for which a procrastination score variable was entered, i.e., the sum of the 14 questions. The minimum score possible on the PS14 item scale is 14 and the maximum score is 56.

Learning style instrument
Kolb’s (1981) measure is a twelve-sentence completion list. Scores are self-ranked on a 4-point, Likert-type scale that has a test, re-test reliability of .92 to .97 (Veres, Sims, & Locklear, 1991). The learning style survey is an assignment in the text used in the public speaking class (Sellnow, 1999b). Students rank their learning style preferences of Assimilator (RO/watch), Converger (AC/think), Accommodator (AE/do), or Diverger (CE/feel) by selecting “1, Least like me,” “2, Third most like me,” “3, Second most like me,” or “4, Most like me.” The four individual learning style scores range from 12 (“Least like me”) to 48 (“Most like me”). The four respective scores always total 120.

Each self-ranked total was entered as a variable: feel (CE), watch (AO), think (AC), and do (AE), as well as each student’s primary learning style type, which was determined by subtracting the CE score from the AC score and the RO score from the AE score. These two differences were plotted on Kolb’s (1981) “Learning Style-Type Grid” and determined the participants’ primary learning type. This procedure was repeated with the difference between the total means to determine the participants-as-a-group primary type. To determine the group’s learning style profile, learning style score means were plotted on Kolb’s “Cycle of Learning Graph.” Profiles by sex were not plotted.

To facilitate the comparison of procrastination tendencies between learning style preferences, scores were summed to create various combination variables. These variables were (1) the CE/AC (feel/think) dimension, (2) the AE/RO (do/watch) dimension, (3) the CE/AE (feel/do) types, and (4) the RO/AC (watch/think) types. Lastly, dichotomous coding of the highest score of each feel/do and watch/think types created a variable that encompassed all learning style preferences. In total, there were nine learning style variables for testing against two procrastination variables.

Distribution analysis
Each learning style and procrastination variable was tested for normal distribution by examining the ratio of skew to standard error, the 95 percent confidence interval ranges, and the Kolmogorov-Simirnov with Lilliefors Significance Correction tests of normality. It was determined to use log (base 10) transformation if any variables were a probability of more than .05 on the tests of normality; this transformation established a more normal distribution of scores with z scores. However, two variables exceeded .05, and all variables had outliers after log transformation; therefore, it was decided to eliminate all outliers. Elimination of outliers in one variable inevitably created an outlier in another variable that did not have one before. These occurrences supported Kolb’s (1984) theory that learning style scores are dialectically opposed rather than unitary. Knowing that outliers would most likely not be eliminated, it was decided to use the first log transformation variables for hypothesis testing and measures of association.

Hypothesis testing
To determine if procrastination by learning style preference would be attributed to a normal distribution in the population, one-way ANOVA procedures were used to the test TPS16 point and PS14 point procrastination means between and within each learning style variable. Significant results, p < .05, were hypothesized for watch (AO), think (AC), summed RO/AC (watch/think) types, and coded CE/AE RO/AC (feel/do, watch/think) types. Individual procrastination questions were also tested with one-way ANOVA procedures to check what types of self-talk occurred with each learning style. There was no prediction as to which questions would be significant for each type. All significant results were tested further with Bonferroni and Tukey a methods, to identify which specific means differ.

Measures of association
To determine if procrastination and learning style preference were related, Pearson correlation coefficients were used to measure the strength of the association between the variables. Positive values for a variable indicated that as procrastination increases, so does the strength of the learning style preference. A negative value indicated that, as procrastination increases, the strength of the learning style preference decreases. For example, if, as hypothesized, the CE/AE (feel/do) preference procrastinates less than the RO/AC (watch/think), the correlation coefficients would be negative CE/AE (feel/do) and positive RO/AC (watch/think).

Distribution of participants was reflective of the usual enrollment in this university’s public speaking classes. Specifically, there were 92 first-year students, 53 sophomores, 14 juniors, and 7 seniors; 101 were male and 65 were female students. The highest percentage of students was first-year, 55.4 percent, and sophomore, 31.9 percent; years with lower representation were juniors, 8.4 percent, and seniors, 4.2 percent, N=166. Dstributiion between sexes was 39.2 percent female, n =65, and 60.8 percent male, n=101.
Factor analysis of the 35-item procrastination scale failed to duplicate Tuckman’s original 16-item scale (Table One). Instead, using principal component analysis, factor analysis of the 35-item procrastination scale extracted 11 components. Component 1 had an eigenvalue of 7.9 and explained 22 percent of the variance. Components 2 through 11 had eigenvalues ranging from 6.6 to 2.9, with all 11 components explaining 63% of the variance. The remaining components (12-35) had eigenvalues ranging from .98 to .21. Using Tuckman’s (1991) original selection criteria of a .40 loading, Component 1 yielded a 1-factor rotated component matrix solution of 14 items with an alpha reliability of .86. Cronbach alpha reliability of the 35-item scale was .79 and of the 16-item scale (TPS16) was .88. The 35-item scale was not used in any further procedures due to reliability less than .80; instead, the TPS16 and the new scale from this study’s factor analysis, PS14, were used.

Scales’ 16 and 14 procrastination score means were midway between “that’s my tendency” and “that’s not my tendency, ” with respective means of 39.64 and 34.34 (SDs = 7.25 and 6.32, respectively; N=16). Tests of normality pre-log (base 10) transformation significances were .20 each. Post-log transformation significances were .01 and .05, respectively. In other words, conversion of procrastination scores resulted in a more normal procrastination distribution, thus increasing test results reliability.
Learning style primary types were evenly distributed for all participants (Table 2). While the individual learning style preference standard deviations were larger than for a normal distribution, the ratio of skew to standard error of the mean indicated that distribution is symmetric (Table 2). Learning style score ranges for Divergers (CE/feel), Convergers (AC/think), and Accommodators (AE/do) met the 95 percent confidence interval. Scores for Assimilators (RO/watch) had the most variance, with 33 percent of the scores below the 95 percent confidence interval range of 27 to 45. Tests of normality were df 166, sig. = .04, .05, .06, and .04, respectively. A log (base 10) transformation yielded normality tests with significance of .00, .05, .00, and .01, respectively.

Plotting learning style score means on Kolb’s (1981) “Cycle of Learning Graph” reflected a group learning style profile of Assimilator (RO/watch), Converger (AC/think), Accommodator (AE/do), and Diverger (CE/feel) with a primary learning type of Assimilator. The data points were very close to the grid center, indicating that, while participants prefer abstract learning through reflective observation, participants had a balanced approach in their learning style. Repeating the same procedure for each sex resulted in a Diverger (CE/feel) type preference for female students and an Assimilator (RO/watch) type preference for male students. While female students preferred concrete learning through action after reflection, male students preferred abstract learning through reflective observation. Again, data points’ placements on the grid reflect a balanced approach in learning.

Combined learning style dimensions and types’ standard deviations were also larger than a normal distribution (Table 3); however, the ratio of skew to standard error of the mean indicated that the distribution was symmetric. Ranges for the CE/AC (feel/think) dimension, the AE/RO (do/watch) dimension, and the CE/AE (feel/do) combined types met the 95% confidence interval. Scores for RO/AC (watch/think) combined types had the most variance, with 5% of the scores placing below and 1% above the 95% confidence interval range of 45 to 77. Tests of normality were df 166, sig. = .20. Log (base 10) transformation of the four variables, in the above order, yielded normality test significances of .07, .01, .20, and .02. Normality test significances for the Feel/Do Watch/Think factor were .00 and .20 pre and post log transformation.

Results: Hypothesis One
There was limited support of Hypothesis One. Findings indicate that levels of procrastination between learning style preferences are distinctly different for males and support Kolb’s (1984) learning model that consists of two dimensions with polar opposites (Table 4). For male students only, one-way ANOVA procedures by learning style dimensions were significant, p = .03, and explain 4 percent of the procrastination variance. A Type I error, the null hypothesis may be true or that procrastination may not be predicted by learning style, cannot be ruled out based on homogeneous subsets of Tukey a post hoc tests. All other one-way procedures were not statistically significant.

Results: Hypothesis Two
Procrastination order on both scales was Assimilator (RO/watch), Converger (AC/think), Accommodator (AE/do), and Diverger (CE/feel). This order failed to support Hypothesis Two, that procrastination rank order, from highest to lowest, is Assimilator, Converger, Diverger, and Accommodator. In order, the means for learning style preference procrastination were TPS16: 37.96, 39.56, 40.80, and 41.11 (SDs = 7.23, 7.16, 7.84, and 6.47, respectively) and PS14: 32.93, 34.39, 35.14, and 35.63 (SDs = 6.34, 5.88, 35.14, and 6.01, respectively). The combined types procrastination order is the reverse of what was hypothesized; rank procrastination order was CE/AE (feel/do) rather then RO/AC (watch/think). Respective means and standard deviations (in parentheses) were TPS16: 38.7 (6.7) and 41.9 (7.7), and PS14: 33.5 (5.7) and 36.6 (6.8).

Results: self-talk questions
The individual questions of the procrastination scale were examined to determine what type of self-talk is employed when procrastinating. Tests of normality were significant for all procrastination questions (df 166, sig. = .00 for each). Significant, p < .05, one-way ANOVA procedures for questions on the procrastination scales are shown in Table 5. Questions beginning with a letter are found on the TPS16 scale; questions beginning with a number are found on the PS14 scale. Self-talk for all participants were self-descriptions of low self-efficacy, self-blame, and a tendency to avoid tasks, explaining 2 percent of the procrastination variance.

Female student self-talk appeared to be self-descriptions of low self-efficacy, self and external blame, and a tendency to avoid tasks, explaining 4 percent to 6 percent of the procrastination variance. Self-talk for male student selections was self-descriptions of low self-efficacy, self and external blame, and a tendency to avoid tasks, explaining 3 percent to 4 percent of the procrastination variance.

Results: measures of association
Contrary to Hypothesis One, the summed CE/AE (feel/do) types had a positive correlation to procrastination, and the RO/AC (watch/think) types had a negative correlation (Table 6). However, when learning styles are looked at individually, none had a correlation to procrastination. This association indicates that students preferring the concrete/active (CE/AE) approach to learning are most likely to procrastinate. Male students preferring the Converger (AC/think) learning style had a negative correlation to procrastination. This correlation indicates that male students preferring an abstract (AC) approach to learning are least likely to procrastinate.

Individual questions of the procrastination scale were tested to determine what type of intrapersonal communication or self-talk is related to procrastination. Positive correlations (Table 5) were considered only for questions with a significant ANOVA. Note, all but one significant question is found on both scales. Question (V & 7) “I am a time waster now, and I can’t seem to do anything about it” has a positive correlation for all students favoring Accommodator (AE/do) and the summed CE/AE (feel/do) types. There was a negative correlation for all students favoring the Assimilator (RO/watch) and Converger (AC/think) learning styles, and summed RO/AC (watch/think) types. This question is a self-description of low self-efficacy and self-blame. These self-descriptions indicate that students who prefer to learn by the concrete and active experimentation approaches (CE/AE) are more inclined to a poor self-image and blame themselves for an inability to integrate courses of action that avoid procrastination. Those students who learn by abstract concepts and reflective observation (RO/AC) have a more positive self-image are less likely to blame themselves for their procrastination (Table 6).

For female students, question (5) “I get right to work, even on life’s unpleasant chores” was positively correlated to the summed CE/AC (feel/think) dimension and negatively correlated to the summed AE/RO (do/watch) dimension. The reverse was true for question (R & 6) “I am an incurable time waster,” which was positively correlated to AE/RO (do/watch) and negatively correlated to CE/AC (feel/think). Question (5) is a self-description of a tendency to avoid tasks, indicating that female students’ preference for learning by concrete experience and abstract concepts (CE/AC) results in task avoidance, whereas female students who prefer active experimentation and reflective observation (AE/RO) learning are less likely to avoid tasks. Question (R & 6) is a self-description of low self-efficacy, indicating that female students who favor concrete experience and abstract learning concepts tend towards negative self-efficacy, whereas those who favor active experimentation/reflective observation learning tend towards positive self-efficacy (Table 6).

For male students, question (CC & 12) “I always finish important jobs with time to spare” and question (HH & 14) “Putting off until tomorrow is not the way I do it” were negatively correlated to Convergers (AC/think). Question (HH & 14) was also negatively correlated to the CE/AC (feel/think) and positively correlated to AE/RO (do/watch). Both questions are self-descriptions of a tendency to avoid tasks. These results indicate that male learning style preference for learning by active experimentation/reflective observation (AE/RO) results in task avoidance, whereas male students who prefer abstract thought (AC/think) and concrete experience/abstract concepts (CE/AC) learning are less likely to avoid tasks.

Questions (V & 7) “I am a time waster now, and I can’t seem to do anything about it” and (FF & 13) “I get stuck in neutral even though I know how important it is to get started” self-described a tendency toward low self-efficacy and self-blame. Correlations were positive for Diverger (CE/feel) and summed CE/AE (feel/do) types, indicating that male students preferring to learn by the concrete and active experimentation approaches (CE/AE) are more inclined to blame themselves for procrastinating behavior with a corresponding poor self-image. Correlations were negative for Assimilator (RO/watch), Convergers (AC/think), and summed RO/AC (watch/think) types, indicating that male students who learn by abstract concepts and reflective observation (RO/AC) are less prone to low self-blame and self-esteem; that is, they are better able to determine a course of action that avoids procrastination with a corresponding positive self-image (Table 6).

Personality makeup
This analysis is the second study to relate similar personality characteristics to academic procrastination. Ferrari et al. (1992) found a positive correlation to Perceiving, one of the four preferences on the MBTI psychological type scale. Characteristics of Perceiving include being spontaneous, open-ended, flexible, adaptable, loose and open to change, and feeling energized by last-minute pressures (Briggs Myers, 1993). Personality characteristics of the summed CE/AE (feel/do) types include empathy, imagination, extroversion, and risk-taking; Accommodator (AE/do) characteristics include adaptation, risk-taking, use of trial and error, and high reliance on others for information (Kolb, 1978, 1984). Notice the similarities between Perceiving and Accommodator; both are impulsive and carefree with Diverger (CE/feel) adding the aspect of sociability. An impulsive and carefree approach to academic tasks can result in time frittered away and deadlines shirked.

On the other hand, Convergers (AC/think) tend to be less sociable and focus on developing practical solutions. These characteristics are similar to Judging, also on the MBTI scale; Judging, as does Converger, has a negative correlation to procrastination (Ferrari et al., 1992). Convergers tend to pre-plan and complete tasks before deadlines; consequently, it appears that the major difference between Converger and Accommodator is the ability to self-regulate performance.
An explanation for this occurrence, may be found in Kolb’s (1976) research on learning style principles: certain careers are associated with certain learning styles; students tend to select a major that best fits with their learning style; and students tend to change majors or learning style preference if there is a mismatch (Kolb, 1978). The university from which data were gathered specializes in engineering and agriculture. Engineering is primarily associated with the Converger (AC/think) learning style and agriculture with Converger and Assimilator (RO/watch) types. Smaller specialties offered are secondary education, business, and arts such as theater and journalism, all of which are associated with either Accommodator (AE/do) or Diverger (CE/feel) types (Kolb, 1981, 1984). Over one-third (41 percent of the learning style types are Accommodators and Divergers in the first and second year grades. Distribution of the Accommodator and Diverger learning styles suggests that students specializing in engineering and agriculture have either not selected a major that best fits with their learning style, or have not changed their major or learning style preference to relieve the inconsistency, whereas students preferring a Converger style have selected a major compatible with their learning style preference. Findings indicate that failure to resolve the mismatch of learning style to major seems to result in procrastination tendencies for Accommodators and Divergers at this university.

Intrapersonal communication
The premise that students favoring Assimilator (RO/watch) and Converger (AC/think) styles are more likely to procrastinate presumes that both styles are prone to negative self-efficacy, which is based on the personality descriptors for Judging, i.e., self-doubter, worrier, feelings of guilt, and doubting one’s own capability to perform (Briggs Myers, 1993; Tuckman, 1991). The assumption that Assimilator and Converger styles have an inability to self-regulate performance is founded on personality preference that these styles would rather generate ideas and defer action. Correlations suggest the opposite; that is, Accommodator (AE/do) and Diverger (CE/feel) styles have negative self-efficacy and less ability to self-regulate performance; Converger styles have positive self-efficacy and self-regulate performance. Evidence of negative self-efficacy is found for all students favoring Accommodator (AE/do) and the summed CE/AC (feel/do) dimension in the negative self-talk and negative self-concept, “I am a time waster now” (Table 7). Furthermore, there is evidence of an inability to integrate a course of action or regulate self-performance in the statement, “and I can’t seem to do anything about it.”

Female students’ Accommodator (AE/do) and Diverger (CE/feel) types express a negative self-concept when they describe themselves as “incurable time wasters.” Male students’ Accommodator and Diverger types express negative self-esteem by describing themselves as “time wasters” and being “stuck in neutral.” In addition, male students differ from female students by describing a tendency to avoid tasks or regulate self-performance. Males depict themselves as “putting off until tomorrow” as the way to “do it.” On the other hand, male Converger (AC/think) types assert, “I always finish important jobs with time to spare” (Table 7).

What may be underlying this negative self-talk, negative self-efficacy and unregulated self-performance is a resentment of structure and rules, a procrastination characteristic of Perceiving (Briggs Myers, 1993). Evidence of structure and rules can be found in questions that positively correlated to Accommodation (AE/do) and Divergers (CE/feel) summed types. Academic structure and rules are apparent in the implied expectations of deadlines, studying, and ability to prioritize. Procrastinators “wait until the last minute,” find studying “boring,” and report inertia even though they “know how important it is to get started.” Perceived negative societal attitudes towards procrastination are apparent in the implication that procrastination is a disease or malady as procrastinators self-describe themselves as “incurable.” Negative self-talk and general self-efficacy are strong predictors of procrastination, thus an area for future research is to determine if there is a direct relationship between learning style preference and self-efficacy (Ferrari et al., 1992; Knaus, 1995).

Whereas examination of individual procrastination scale questions provides us with a sense of the kind of self-communication occurring, there has been little research on procrastination as a communication strategy. Behnke and Sawyer (1999) have correlated procrastination to communication apprehension while Chissom and Iran-Nejad (1992) have specifically identified procrastination as a strategy students use to avoid learning. An area of promising research identifies procrastination as a compliance-gaining strategy (Schneider & Beaubien, 1996). Typically, procrastination is viewed as avoidance of an interpersonal interaction; however, Schneider and Beaubien describe it as an intrapersonal communication device employed by physicians to avoid a difficult (i.e., no-win) situation. In a compliance-gaining situation, an agent selects a strategy (intrapersonal communication), which results in a persuasive interaction with a target (interpersonal communication) and (hopefully) the desired outcome for the agent. If procrastination is selected as the compliance-gaining strategy (intrapersonal communication), interaction avoidance and (hopefully) the desired outcome for the agent result.

Applying this framework suggests that persuasive interaction (interpersonal communication) and procrastination (intrapersonal communication) are subject to situational factors and the level of power the agent and target (deTurck, 1995). For example, students often complete major papers the night before they are due. A student’s decision (intrapersonal communication) to complete a major assignment the night before it is due (procrastination strategy) may be predicated on the current relationship (referent power) with the target or experiences (situational factors) that resulted in past high grades. The target may accept this interaction avoidance because an instructor prefers not to use his/her authority to discipline or otherwise control (coercive, legitimate, and expert power) the agent’s behavior. Eventually, this strategy may become habit on the part of the agent and culturally accepted on the part of the target (Hill, Hill, & Barrall, 1978; Knaus, 1995). Findings from this study indicate that the agent is most likely a risk-taking Accommodator (AE/do) that feels energized by last-minute pressures. Further research on procrastination as a compliance-gaining technique, as opposed to a personality characteristic, is warranted.

Tuckman’s (1991) procrastination survey was devised to help college students “regulate their own learning.” (p. 474). Due to its high reliability, .88, the 16-point scale is an excellent tool for student self-administration, self-evaluation, and self-use; making use of it in college student class advising and planning divisions may be useful. However, a limitation of the scale is that some students found the phrasing puzzling, and as with all self-report tools, results are contingent on the honesty of participants. In this region, another limitation is a “desire to please” which may have influenced some responses. In addition, two methods were employed to collect these data: completion in the class and take home. Not all the take-home surveys were returned, leading to the conclusion that some high procrastination scores are missing. The sample itself, which may have been underrepresented with juniors and seniors, was primarily white, Midwestern students whose procrastination tendencies may not be representative of other regions in the United States.

Other findings that may be attributed to regional differences are a factor analysis of 14 items instead of 16 on the 35-item procrastination scale. The 14-item scale suggested self-blame, whereas the 16-item scale includes external blame for procrastination tendencies. The above findings suggest that researchers may want to include factor analysis and alpha reliability when using the 35-item scale in order to assure a scale reliability of at least .80 and to ascertain regional differences. Lastly, while the transformed raw data from individual learning types and procrastination indicated a normal population distribution, there is less confidence in the transformed summed dimensions and types. This limitation should be kept in mind when applying results.

The general purpose of this study was to determine if procrastination could be predicted by known personality characteristics such as learning style preference. Although there is some evidence to support this association, limitations indicate that caution is needed with using this information. A more specific purpose was to introduce the investigation of procrastination to the field of communication. To date, research has been limited, yet it is evident from this study that content analysis of significant survey questions can be interpreted as self-talk, providing a basis for qualitative exploration during focus group and interview sessions. For example, society may view procrastination as a disease, and students agree with this view by describing themselves as “incurable” when answering a survey, but will they support this theme during a qualitative process in which they describe their internal perspective of procrastination?

Another area of investigation is within intrapersonal communication, specifically, as a compliance-gaining strategy used to convert no-win situations to ones of “informed delay” (Ferrari et al., 1995, p. 4). The earliest use of procrastination had one definition, an “informed delay,” and was thought to be a wise or prudent choice (p. 4). However, delay may be an unavoidable choice for some students as today’s college students are busy people. Within a semester’s 15-week time span, students carrying full academic loads are attending classes 12 or more hours a week, with a corresponding homework load of 36 hours or more. Many are also involved in extracurricular activities, such as sports and other campus organizations, which can vary from an occasional hour to 10 hours or more a week in active participation. Many more must work a part-time job, usually 20 hours a week, to help finance their college opportunity. Given this academic lifestyle, it is surprising that academic procrastination rates are not higher. Perhaps it is time to include informed delay within the definition of academic procrastination.

Lastly, expanding the understanding of procrastination by way of a communication lens may augment the multitude of self-help resources advocating the use of self-talk as a “cure” for procrastination. As a behavior, procrastination is obvious, however, as an intrapersonal communication device, procrastination is less clear; this study offers new insight into procrastination as a strategy one employs after self-talk has occurred, thus inviting further research through a communication standpoint.


Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall.
Behnke, R., & Sawyer, C. (1999, Fall). Public speaking procrastination as a correlate of public speaking communication apprehension and self-perceived public speaking competence. Communication Research Reports, 16 (1), 40-47.
Briggs Myers, I. (1993). Introduction to type (5th ed.). (Rev. by L. K. Kirby & K. D. Myers). Palo Alto, CA: Consulting Psychologists Press.
Chissom, B., & Iran-Nejad, A. (1992). Development of an instrument to assess learning strategies. Psychological Reports, 71, 1001-1002.
Curry, L. (1987). Integrating concepts of cognitive or learning style: A review with attention to psychometric standards. Ottawa, Ontario: Canadian College of Health Service Executives.
DeTurck, M. (1995). Persuasion and compliance-gaining. In R. Hartman & L. Texter (Eds.), Advanced interpersonal communication (pp. 271-284). Dubuque, IA: Kendall/Hunt.
Ferrari, J., & Beck, B. (1999, Summer). Affective responses before and after fraudulent excuses by academic procrastinators. Education, 118 (4), 529-538.
Ferrari, J., Johnson, J., & McCown, W. (1995). Procrastination and task avoidance theory, research and treatment. New York: Plenum Press.
Ferrari J., Parker, J., & Ware, C. (1992). Academic procrastination: Personality correlates with Myers-Briggs types, self-efficacy, and academic locus of control. Journal of Social Behavior and Personality, 7 (3), 495-502.
Haycock, L., McCarthy, P., & Skay, C. (1998, Summer). Procrastination in college students: The role of self-efficacy and anxiety. Journal of Counseling and Development, 76, 317-324.
Hickcox, L. (1995). Learning styles: A survey of adult learning style inventory models. In R. Sims & S. Sims (Eds.), The importance of learning styles (pp. 25-47). Westport, CT: Greenwood Press.
Hill, M., Hill, D., Chabot, A., & Barrall, J. (1978). A survey of college faculty and student procrastination. College Student Journal, 12, 256-262.
Knaus, W. (1995). Overcoming procrastination: A new look [On-line]. Retrieved from Albert Ellis Institute Home Page World Wide Web: Accessed January 14, 2000.
Kolb, D. A. (1976). Learning style inventory technical manual. Boston: McBer and Company.
Kolb, D. A. (1978). Learning style inventory technical manual (Rev. ed.). Boston: McBer and Company.
Kolb, D. A. (1981). Learning style inventory: Self-scoring inventory and interpretation booklet. Boston: McBer and Company.
Kolb, D. A. (1984). Experiential learning experience as the source of learning and development. Englewood Cliffs, NJ: Prentice-Hall.
Kolb, D. A., Rubin I., & McIntyre, J. (1971). Organizational psychology an experiential approach. Englewood Cliffs, NJ: Prentice-Hall.
Lay, C. (1986). At last, my research article on procrastination. Journal of Research in Personality, 20, 474-495.
Milgram, N. (1995, March). Correlates of academic procrastination: Discomfort, task aversiveness and task capability. The Journal of Psychology, 129 (2), 145-156.
Roberts, C., Edwards, R., & Barker, L. (1987). Intrapersonal communication processes. Scottsdale, AZ: Gorsuch Scarisbrick.
Roig, M., & DeTommaso, L. (1995, October). Are college cheating and plagiarism related to academic procrastination? Psychological Reports, 77 (2), 690-698. Abstract from…1/te%20%20procrastination/di%200034 Accessed January 14, 2000.
Schneider, D., & Beaubien, R. (1996, Summer). A naturalistic investigation of compliance-gaining strategies employed by doctors in medical interviews. The Southern Communication Journal, 61 (4), 332-341.
Sellnow, D. (1999a). The process of public speaking (Rev. ed.). Orlando, FL: Harcourt College Custom Publishers.
Sellnow, D. (1999b). Public speaking applications (2nd ed.). New York: Forbes Custom Publishing.
Senecal, C. (1995, October). Self-regulation and academic procrastination. The Journal of Social Psychology, 135 (5), 607-620.
Solomon, L., & Rothblum, E. (1984). Academic procrastination: Frequency and cognitive-behavioral correlates. Journal of Counseling Psychology, 31 (4), 503-509.
Tuckman, B. (1989, March). Procrastination busting. Educational Technology, 29 (3), 48-49.
Tuckman, B. (1991). The development and concurrent validity of the procrastination scale. Educational and Psychological Measurement, 51, 473-480.
Tuckman, B. (1998, Winter). Using tests as an incentive to motivate procrastinators. The Journal of Experimental Education, 66 (2), 141-148.
Veres III, J., Sims, R., & Locklear, T. (1991). Improving the reliability of Kolb’s revised learning style inventory. Educational and Psychological Measurement, 51, 143-150.

* Denise Elmer is an instructor in the Department of Communication, Drama and Journalism at Angelo State University, San Angelo, TX, 76909. She may be contacted at her personal e-mail address <>. Portions of this research were presented at Central States Communication Association convention April 2000, Detroit, MI. The author thanks the two anonymous reviewers and Dr. R. Collins for their assistance in the development of this manuscript.