Available in: http://www.redalyc.org/articulo.oa?id=17217376006

Scientific Information SystemNetwork of Scientific Journals from Latin America, the Caribbean, Spain and

Portugal

Sabry M. Abd-El-Fattah

Garrison's Model of Self-Directed Learning: Preliminary Validation and Relationship to

Academic Achievement The Spanish Journal of Psychology, vol. 13, núm. 2, 2010, pp.

586-596,

Universidad Complutense de Madrid

España

How to cite Complete issue More information about this article Journal's homepage

The Spanish Journal of Psychology,

ISSN (Printed Version): 1138-7416

psyjour@sis.ucm.es

Universidad Complutense de Madrid

España

www.redalyc.orgNon-Profit Academic Project, developed under the Open Acces Initiative

586

In this project, 119 undergraduates responded to a questionnaire tapping three psychological constructs implicated in Garrison’s model of self-directed learning: self-management, self-monitoring, and motivation. Mediation analyses showed that these psychological constructs are interrelated and that motivation mediates the relationship between self-management and self-monitoring. Path modeling analyses revealed that self-management and self-monitoring significantly predicted academic achievement over two semesters with self-management being the strongest predictor. Motivation significantly predicted academic achievement over the second semester only. Implications of these findings for self-directed learning and academic achievement in a traditional classroom setting are discussed.Keywords: self-directed learning, self-management, self-monitoring, motivation, path modeling, mediation analysis.

En este trabajo, 119 estudiantes posgraduados fueron evaluados de acuerdo con un cuestionario que

mide tres constructos psicológicos implicados en el modelo de Garrison sobre aprendizaje autodirigido:

autogestión, autoseguimiento, y motivación. Los análisis de mediación mostraron que estos constructos

psicológicos están interrelacionados y que la motivación media en la relación entre autogestión y

autoseguimiento. Los análisis de Path modeling indicaron que la autogestión y el autoseguimiento

predecían significativamente el logro académico a lo largo de dos semestres en los que la autogestión

era el mejor indicador. La motivación predecía significativamente el éxito académico sólo a lo largo

del segundo semestre. Se discuten las implicaciones que estos resultados pueden tener sobre el

aprendizaje autodirigido y el éxito académico en el contexto de un aula tradicional.

Palabras clave: aprendizaje autodirigido, autogestión, autoseguimiento, motivación, Path modeling,

análisis de mediación.

Garrison’s Model of Self-Directed Learning: Preliminary Validation and

Relationship to Academic Achievement

Sabry M. Abd-El-Fattah

Minia University (Egypt)

Correspondence concerning this article should be addressed to Sabry M. Abd-El-Fattah. Department of Educational Psychology.Faculty of Education. Minia University. Minia. (Egypt). E-mail: sabryrahma@hotmail.com

Copyright 2010 by The Spanish Journal of PsychologyISSN 1138-7416

The Spanish Journal of Psychology 2010, Vol. 13 No. 2, 586-596

GARRISON’S MODEL OF SELF-DIRECTED LEARNING 587

Self-directed learning (SDL) is a central concept in the study and practice of adult education (Brockett & Hiemstra, 1991). As important as the construct is to adult education, little attention has been directed to the cognitive and motivational factors implicated in SDL (Brookfield, 1986; Knowles, 1975). Long (1989, p. 3) stressed the importance of the psychological perspective in SDL and argued that “The critical dimension in self-directed learning is not the sociological variable, nor is it the pedagogical factor. The main distinction is the psychological variable.” Furthermore, the overriding theme of SDL has mainly been the external management of the learning process. According to this perspective, the learner exercises a great deal of independence in deciding what is worthwhile to learn and how to approach the learning tasks, regardless of entering competencies and contextual contingencies (Brookfield, 1986). As such, SDL has largely been defined in terms of external control and facilitation, rather than internal cognitive processing and learning.

To address these concerns, Garrison (1997) proposed a SDL model which integrated external management (contextual control), internal monitoring (cognitive responsibility), and motivational (entering and task) factors associated with learning in an educational context. Garrison defined SDL as an approach where learners are motivated to assume personal responsibility and collaborative control of the cognitive (self-monitoring) and contextual (self-management) processes in constructing and confirming meaningful and worthwhile learning outcomes. Garrison’s collaborative perspective of SDL has the learner taking responsibility for constructing meaning (cognitive perspective) while including the participation of others in confirming worthwhile knowledge (social perspective).

According to Garrison (1997), self-management concerns task control issues including the enactment of learning goals and the management of learning resources and support. The essence of the term can be found in the self-regulated motivational literature (Pintrich & DeGroot, 1990). Task control is determined by balancing the factors of proficiency, resources, and interdependence (Garrison, 1993). Proficiency represents the abilities and skills of the facilitator and the learner. Resources encompass a range of support and assistance available in an educational setting. Interdependence reflects institutional or subject norms and standards as well as a learner’s integrity and choice. Self-management of learning tasks represents a collaborative experience between the teacher and the learner. The teacher maintains an appropriate dynamic balance of external control necessary for successful educational outcomes (Prawat, 1992; Resnick, 1991).

Another component of Garrison’s SDL model is self-monitoring. It addresses the cognitive and metacognitive processes which include monitoring the repertoire of learning strategies as well as awareness and an ability to think about our thinking. It is the process whereby the

learner takes responsibility for the construction of personal meaning through integrating new ideas and concepts with previous knowledge. It facilitates a metacognitive perspective on learning and a generalized ability to learn reflectively. Reflective learning can help “develop learners who are capable of monitoring themselves in a variety of situations” (Candy, Harri-Augstein, &Thomas, 1985, p. 115). The teacher can provide effective feedback to help the learner self-monitoring the quality (meaning and validity) of the learning outcome because internal feedback alone may lack accuracy and explicitness. The degree of self-monitoring will depend upon the learner’s proficiency (abilities and strategies) in conjunction with contextual and epistemological demands (Butler & Winne, 1995; Garrison, 1991).

Lastly, motivation helps initiate and maintain effort towards learning and the achievement of cognitive goals. Motivation includes entering motivation and task motivation. Entering motivation establishes commitment to a particular goal and the intent to act. It can be perceived as “commitment--the coming together of attitudes, feelings, and goals” (Thompson, 1992, p. 103). Task motivation is the tendency to focus on and persist in learning activities and goals. As Corno (1989) suggested, “Motivational factors . . . shape intentions and fuel task involvement” (pp. 114-115). It is hypothesized that entering motivation is determined by valence and expectancy. Valence reflects the attraction to particular learning goals. The factors that determine valence are personal needs (values) and affective states (preferences). In a learning context, expectancy refers to the belief that a desired outcome can be achieved (Pintrich & DeGroot, 1990). Task motivation is integrally connected to task control, self-management, and the issue of volition. In a learning context, “volition refers to bringing discordant affective and executional preferences in line with one’s task goals” (Kanfer, 1989, p. 381). Volition is concerned with sustaining intentional effort or diligence that can influence persistence and task performance (Pintrich & DeGroot, 1990).

The literature on SDL asserts that self-directed learners demonstrate great awareness of their responsibility in making learning meaningful and monitoring themselves (Garrison, 1997). They were found to be curious and willing to try new things, view problems as challenges, desire change, and enjoy learning (Temple & Rodero, 1995). Guthrie, McGough, Bennett, and Rice (1996) reported that in a Concept-Oriented Reading Instruction (CORI) program, self-directed learners demonstrated the ability to search for information in multiple texts, employ different strategies to achieve goals, and to represent ideas in different forms (drawing and writing). Morrow and Young (1997) observed that with proper planning and implementation, self-directed learning can encourage students to develop their own rules and leadership patterns. Thus, self-directed learners take responsibility for the construction

ABD-EL-FATTAH588

of personal meaning, initiation and maintenance of effort toward learning tasks, and achievement of cognitive goals, and therefore they are more likely to be promoted as high achievers (Garrison, 1997).

Linking self-directed learning to academic achievement

For example, several studies have reported a significant positive relationship between SDL and academic achievement in a traditional classroom setting (Darmayanti, 1994), a non-web based distance learning setting (Hsu & Shiue, 2005), a web-based learning setting (Haron, 2003), and a distant education setting (Harriman, 1990). Other studies have reported a significant positive relationship within a specific content areas including nursing (Savoie, 1979), social and political sciences (Anderson, 1993), business (Morris, 1995), business, communication, public administration, and hospitality management (Ogazon, 1995), and biology (Haggerty, 2000). In a recent study, Stewart (2007), for example, found a positive relationship between SDL readiness and the overall learning outcome ratings in project-based learning in engineering with self-management being the strongest predictor. Self-control and desire for learning significantly predicted the overall learning outcome ratings but their R2 values were relatively low.

Rational of the study

Self-monitoring is intimately linked to the external management of reaming tasks and activities. An interesting and important issue arises with regard to responsibility (self-monitoring) and control (self-management) within a learning context is whether responsibility must precede control or vice versa. Although theoretically they go hand in hand, it is very difficult for learners to assume responsibility for their own learning without feeling they have some control over the educational transaction. Without choice and collaboration, it may well be unrealistic to expect students to assume responsibility for their learning.

Meanwhile, absolute learners’ control may adversely affect or reduce the efficiency of achieving quality learning outcomes. There is research evidence that collaborative control results in more effective self- monitoring and, therefore improved performance. In addition, sharing control of learning activities and tasks provides opportunities for instructional support while encouraging students to assume cognitive responsibility (Butler & Winne, 1995).

The inseparability of monitoring and managing of the learning process is further complicated by motivational concerns. According to Garrison (1997, p. 9), “Motivation reflects perceived value and anticipated success of learning goals at the time learning is initiated and mediates between context (control) and cognition (responsibility) during

the learning process.” Anticipated control is an important perception when assessing expectancy of success and making decisions regarding goal-directed behavior. It is believed that control expectations “influence the direction of much of our behavior; they help to determine where we invest our achievement energies” (Weisz, 1983, p. 234). If students are to have an expectation of control, they must have some choice over their educational goals. Providing opportunities for control and choice from the beginning can significantly strengthen the entering motivational state, which subsequently influences whether students will take responsibility to be self-directed and persist in their learning tasks.

Although the linkages among the psychological traits implicated in Garrison’s SDL model may be theoretically meaningful and conceptually sound, such relationships have not been empirically validated using real world data. In addition, most of the studies on SDL have been conducted using Western samples (Garrison, 1997). It is, therefore, imperative to examine the practical implications of Garrison’s SDL model in Non-western contexts to determine the generalizability of the findings reported using Western samples. Actually, it is possible that careful examination and validation of Garrison’ SDL model in Non-western contexts bring about profitable use of the model and its applications. On the other hand, uncritical use of Garrison’s SDL model in Non-western contexts to make students more self-directed learners may not necessarily result in better academic outcomes, since countries may differ in values and beliefs about education and opportunities offered to students.

Thus, one goal of the present study is to examine the relationships among self-management, self-monitoring, and motivational factors, associated with learning in an educational context. Of particular interest is to investigate whether motivational factors may mediate the relationship between self-management and self-monitoring. A second goal of the present study is to examine the relationships of these psychological constructs to academic achievement measured over two semesters in a traditional classroom setting. Of particular interest is to relate academic achievement measured over two semesters to the earlier collected psychological indices.

Methods

Participants

Subjects of the present study included 119 first year undergraduates enrolled in an education program in a public university in Minia, Egypt. There were 65 males and 54 females. Fifty-six students majored in science and 63 majored in arts. The mean age of the participants was 18.7 years (SD = 1.7) with a range from 18 to 21 years.

GARRISON’S MODEL OF SELF-DIRECTED LEARNING 589

Measurements

The Self-Directed Learning Aptitude Scale. The instrument most widely used in educational research to measure SDL readiness is Guglielmino’s (1977) Self-Directed Learning Readiness Scale (SDLRS). Based on problems with validity testing of the SDLRS, Field (1989) suggested discontinuing this tool. Furthermore, several studies have raised questions about the reliability of the SDLRS when used in different racial and class populations (Straka 1995). Even though scales such as the SDLRS have been developed, they do not translate easily and are not readily available and incur a cost for their use. Thus, the development of a new scale should allow for these problems to be addressed.

The Self-Directed Learning Aptitude Scale (SDLAS) was developed for the purpose of the present study. The initial item pool comprised 40 items designed to measure students’ aptitude to SDL based on a review of the research of Knowles (1975), Brookfield (1986), and Garrison (1991, 1993). Using a Delphi technique (Sharicey & Sharpies, 2001), the members of an expert panel independently assessed each of the 40 items using a 5-point scale. A score of 1 denoted Absolutely Inappropriate and a score of 5 denoted Absolutely Appropriate. Panel members were given space to modify each item if they chose. Following the guidelines suggested by Kline (1986) and Crocker and Algina (1986), an item was retained when the panel consensus for that item was of at least 80% of appropriateness. Items where consensus was not achieved, but where less than 20% of the panel denoted Inappropriate (i.e. 80% either Appropriate, Absolutely Appropriate, or were Unsure) were retained for the subsequent round. A total of 28 items were retained after two rounds of the Delphi technique.

Procedures Students were recruited to participate voluntary during

their normal classes at their university. All participants were informed that they could deny their participation in data collection without any explanation, penalty, or cost to them. The initial 28 items of the SDLAS were administered to the sample of the study during the third week of the 2008/2009 academic year. Students were asked to describe themselves by indicating on a 4-point Likert-type scale the extent to which each item was descriptive of their own characteristics with a score of 1 denoted Strongly Disagree and a sore of 4 denoted Strongly Agree. In addition, students’ academic achievement scores over two semesters were obtained, with permission, from the university records. These scores were the courses aggregated total score (i.e., the sum of on-courses assignments and examinations score).

Results

Exploratory factor analysis

An exploratory factor analysis with principal components was conducted to identify a viable factor structure of the 28-item pool of the SDLAS. The resulting factors were rotated to a simple structure using oblimin rotation. The number of factors retained was determined using the following criteria: (1) Kaiser’s rule of retaining factors with eigenvalues greater than 1, (2) Cattell’s Scree test of the plot of eigenvalues, and (3) each factor had to have at least three items. Inclusion criteria for items on the retained factor were that they had loadings of at least .30 on that factor. Items with high cross-loadings, wherein an item had a loading of .30 or greater on more than one factor, were assigned to a factor on the basis of logical fit. A corrected item-total correlation of .30 or above was required to confirm the assignment decision. The factors that were identified were named on the basis of their content (Nunnally, 1978).

The analysis retained three factors: self-management (8 items), motivation (9 items), and self-monitoring (9 items). These three factors explained 22, 18, and 15 per cent of the total variance extracted respectively. The loadings of two items, “I enjoy studying” and “I would like to make decisions for myself” fall substantially below the traditional accepted criteria of .30 for retained item loadings (Velicer & Jackson, 1990) and therefore were dropped from the scale. The 26 items of the SDLAS and their loadings on corresponding factors are shown in Table 1. There were no mean differences in self-management, motivation, or self-monitoring, reported in Table 2, due to the effect of gender or academic major (p > .05).

Path modeling

The relationships among self-management, motivation, self-monitoring, and academic achievement over two semesters were analyzed through path modeling, using the partial least squares (PLS) method. This analytical procedure was chosen because it does not hinge upon large samples, it can be used with both formative and reflective constructs, and it does not make assumptions about the underlying data distribution when estimating the model parameters (Chin & Newsted, 1999; Ringle, Wende, & Will, 2005). The software used was SmartPLS 2.0 (Ringle et al., 2005).

In practice, a PLS model is developed in two stages. In the first stage, the measurement model is tested by performing reliability and discriminative validity analyses on each of the measures to ensure that reliable and valid measures of the constructs are being employed. In the second stage, the inner or structural model is tested by estimating the paths between the constructs, determining their significance as well as the predictive ability of the model. The PLS procedure calculates an estimate for each

ABD-EL-FATTAH590

construct or latent variable derived from corresponding observed variables and thus partitioning the hypothesized inner model into its component constructs. To evaluate the model against observed data, an iterative procedure fits observed measures to corresponding latent variables, then estimates the relationships between the latent variables. A least squares fit between observed and modeled parameters is computed. A best-fit solution is regarded as found when the least squares function stabilizes between iterations (Ringle et al., 2005).

Since the approach is variance-based (as distinct from covariance-based), the PLS procedure has been described as soft modeling, and is claimed to be more useful in investigating descriptive and predictive relationships than confirmatory analysis (Sellin & Keeves, 1997). A readable review of the procedure has been published by Haenlein and Kaplan (2004). Two recent examples of researchers using the PLS method within behavioral research are Abd-El-Fattah (2005) and Boman, Smith, and Curtis (2003).

Table 1Oblimin factor loadings, PLS loadings, T values, average variance extracted (AVE), and composite and Cronbach’s alpha reliability coefficients of the Self-Directed Learning Aptitude Scale

Factor/Statistic Factor loadings

PLS loadings

Tvalues* AVE Composite

reliabilityCronbach’s

alpha

Self-management .73 .75 .82

1. I am well-organized in my learning. .83 .81 8.38. I set up strict timeframes to learn something new. .80 .77 6.714. I have good management skills. .75 .68 10.37. I set up planned solutions to solve my problems. .74 .73 9.43. I can decide about the priority of my work. .73 .72 11.32. I can manage pursuing my own learning. .72 .67 5.55. I prefer to plan my own learning. .70 .65 6.418. I am efficient in managing my time. .68 .60 8.3

Motivation .70 .80 .84

2. I take the challenge to learn. .80 .76 7.615. I am a ‘why’ person. .78 .80 10.811. I critically evaluate new ideas and knowledge. .75 .73 5.622. I would like to evaluate the level of my learning progress. .71 .65 4.926. I would like to learn from my mistakes. .67 .77 7.89. I believe in effort to improve my performance. .66 .63 5.61. I enjoy learning new things. .65 .62 4.817. I trust my abilities to learn new things. .64 .73 8.312. I have positive expectations about what I am learning. .62 .75 10.7

Self-monitoring .68 .81 .86

6. I am a ware of my own weaknesses. .81 .76 12.74. I can link pieces of information when I am learning. .80 .77 8.513. I pay attention to all details before taking a decision. .77 .73 4.719. I would like to set up my goals. .75 .72 6.323. I correct myself when I make mistakes. .71 .80 8.925. I am a responsible person. .69 .74 10.421. I judge my abilities fairly. .66 .73 7.524. I think deeply when solving a problem. .65 .68 6.616. I prefer to set up my criteria to evaluate my performance. .63 .70 10.3

Note. N = 119. *T-values are significant for all instances, p < .01.

GARRISON’S MODEL OF SELF-DIRECTED LEARNING 591

The assessment of the measurement model

In the PLS method, a reflective measurement model is assessed by evaluating: (a) the reliability of single items, (b) the convergent validity of a construct, and (c) the discriminant validity of a construct (Hulland 1999).

For the assessment of individual item reliability, a rule of thumb recommends accepting items with loadings of at least .5, particularly when new items or newly developed scales are employed (Chin, 1998; Hulland 1999). Table 1 shows that all items loaded on their corresponding factors at .60 or higher (t > 4.7, p < .01 for all items).

For the assessment of convergent validity (also referred to as internal consistency) of a construct, a rule of thumb recommends accepting a construct with Cronbach’s alpha and/or composite reliability coefficient that exceeds the threshold of .6 as proposed by Bagozzi and Yi (1988), or .7 as suggested by Nunnally (1978). Table 1 shows that both Cronbach’s alpha and the composite reliability coefficients for self-management, motivation, and self-monitoring were > .75. Another index that is used to assess a construct convergent validity is the average variance extracted (AVE). The AVE measures the amount of variance captured by the construct relative to the amount of variance due to measurement error. The AVE should exceed a value of .5 to highlight a construct convergent validity (Fornell & Larcker

1981). Table 1 shows that the AVE for self-management, motivation, and self-monitoring were > .68.

For the assessment of discriminant validity of a construct, a rule of thumb recommends accepting a construct if the square root of the AVE of that construct is larger than the Pearson’s correlation coefficients of that construct to all other latent constructs in the model (Fornell & Larcker, 1981). Unfortunately, guidelines about how much larger the AVE should be than these Pearson’s correlation coefficients are not available (Gefen, Straub, & Boudreau, 2000). Table 3 shows that the square roots of the AVEs for self-management, motivational factors, and self-monitoring are greater than the Pearson’s correlation coefficients of these constructs with each other. Based on the findings, one can concluded that the measurement model fits the present data set adequately.

The assessment of the structural model

To evaluate the specified structural model, displayed in Figure 1, one should report the standardized ß coefficients and t values along with the R² (Chin 1998; Hulland 1999). The standardized ß coefficient is a measure of the strength of the relationship between an independent and a dependent variable while holding constant the effects of the remaining

Table 2Gender and academic major mean differences in self-management, motivation, self-monitoring

Variables /Statistic n M SD F (1,117)

Gender

Self-management 1.3Males 65 23.7 3.4

Females 54 22.9 3.2Motivation .97

Males 65 24.9 3.6Females 54 24.3 3.0

Self-monitoring .83Males 65 26.3 4.2

Females 54 25.7 3.5Major

Self-management 1.4Science 56 21.6 2.9

Arts 63 20.9 2.5Motivation .61

Science 56 25.7 3.8Arts 63 25.3 3.4

Self-monitoring .67Science 56 23.7 4.3

Arts 63 23.2 3.8

Note. N = 119. df = 1, 117. * F-values are nonsignificant for all instances, p > .05.

ABD-EL-FATTAH592

independent variables. The R² represents the proportion of variance on a dependent variable explained by all variables jointly. A t statistics is used to examine the significance of the parameter estimates (Hulland, 1999). In sum, it was found that self-management significantly predicted all constructs in the model (p < .01). Motivation significantly

predicted self-monitoring and achievement in Semester 1 (p < .01) but not achievement in Semester 2 (p > .05). Self-monitoring significantly predicted academic achievement over the two semesters (p < .01). Finally, academic achievement in Semester1 significantly predicted academic achievement in Semester 2.

Table 3Pearson’s correlation coefficients, the square root of average variance extracted (AVE), means, and standard deviations of self-management, motivation, and self-monitoring

Variables 1 2 3 M SD

1. Self-management .87 .36** .31* 3.2 .742. Motivation .84 .34** 2.8 .643. Self-monitoring .82 3.4 .66

Note. N = 119. Values on diagonal represents the square root of the average variance extracted (AVE) statistic as used within PLS. Means are expressed along a four point type-Likert scale, 1-4, summed across items.*p < .05; **p < .01.

Figure 1. PLS model of the relationship among self-management, motivation, self-monitoring, achievement in Semester 1, and achievement Semester 2 (Dashed line indicates nonsignificant effect)

.81

.77

.68

.73

.72

.67

.65

.60

.76 .80 .73 .65 .77 .63 .62 .73 .75

.76

.77

.73

.72

.80

.74

.73

.68

.70

.29(t = 6.74)

.31 (t = 5.52)

.30 (t = 6.51)

.36(t = 7.62)

.32 (t = 7.53)

.35

(t = 8.42)

.28 (t = 7.53) .12

(t = 1.32)

.41(t = 9.33)

.31

(t=

3.62

) R2 = .09

R2 = .19

R2 = .30 R2 = .40

Motivation

AchievementSemester1

AchievementSemester 2

Item 2 Item15 Item 11 Item 22 Item 26 Item 9 Item 10 Item 17 Item 12

Self-managment

Item 18

Item 5

Item 20

Item 3

Item 7

Item 14

Item 8

Item 1

Self-monitoring

Item 6

Item 4

Item 13

Item 19

Item 23

Item 25

Item 21

Item 24

Item 16

GARRISON’S MODEL OF SELF-DIRECTED LEARNING 593

Mediation analysis

According to Baron and Kenny (1986, p. 1176), a variable may be considered a mediator “to the extent that it accounts for the relation between the predictor and the criterion”. The mediation model tested in the present study conceptually denotes self-management as an independent variable (X), motivational as a mediator variable (M), and self-monitoring as a dependent variable (Y). Gender and academic major were also included as covariates (COVs) to test for their possible partial effect on self-monitoring (Y). The SPSS script developed by Christopher Preacher from the University of Kansas and Andrew Hays from the Ohio State University (Preacher & Hayes, 2004) was used to run the mediation analysis and the results are summarized in Table 4.

Baron and Kenny technique showed that self-management significantly predicted self-monitoring (b = .78, SE = .07, t = 11.14, p < .01), self-management significantly predicted motivation (b = .67, SE =.08, t = 8.38, p < .01), and motivation significantly predicted self-monitoring (b = .74, SE = .07, t = 10.57, p < .01). When motivation was controlled for, self-management significantly predicted self-monitoring (b = .66, SE = .08, t = 8.25, p < .01). Gender (b = .13, SE = .11, t = 1.18) and academic major (b = .17, SE = .10, t = 1.70), as covariates, did not have significant effects on self-monitoring (p > .05).

The Sobel test showed identical findings to that of Baron and Kenny technique (effect (z) = .55, SE = .10, z = 5.5, p < .05). The ‘effect’ in Sobel test represents the indirect effect of the mediator. This indirect effect is the product of paths ab and is equivalent to (c - c’) where path c is the total effect and c’ is the direct effect (Sobel, 1982).

The bootstrapping technique (Chernick, 1999) further supported the mediation effect of motivation with a value of zero not included within the lower bound (.26) and the upper bound (.73) of the 95% bias-corrected and accelerated bootstrap confidence intervals. Based on these findings, one can conclude that the motivation mediated the relationship between self-management and self-monitoring.

Discussion

The present analyses revealed several notable findings. First, self-management significantly predicted self-monitoring. This means that increasing a learner’s control through self-management brings with it elevated responsibilities, particularly with regard to the learning process itself and the construction of meaning. The immediate benefit of increased self-management is increased awareness of the need to make learning more meaningful, that is, to take greater responsibility in the monitoring of the learning process itself. It is difficult to get learners to accept responsibility for meaningful learning

outcomes when they have little control of, and input into, the learning process (Garrison, 1991, 1997).

However, as necessary as a sense of control is, without appropriate support and guidance, learners may not persist or achieve the desired educational outcomes. From an educational perspective, quality of learning outcomes is not simply a question of student control and responsibility. The teacher also plays an integral role in education as a transactional process and has his or her own legitimate and often necessary control and responsibility concerns. It is the teacher who is charged with the responsibility of clarifying goals, shaping learning activities, and assessing learning outcomes. Furthermore, in collaboration with students, teachers are responsible for establishing the balance of control by providing support, directions, and standards to ensure worthwhile outcomes and continuing efforts to learn (Garrison, 1993).

The second notable finding concerned motivation as a mediator of the relationship between self-management and self-monitoring. Learners are intrinsically motivated to assume responsibility for constructing meaning and understanding when they have some control over the learning experience. Garrison (1997) theorized that “Motivation reflects perceived value and anticipated success of learning goals at the time learning is initiated and mediates between context (control) and cognition (responsibility) during the learning process.” (p. 9). Garrison (1993) also proposed that “Control and choice strengthen motivation, which in turn builds a sense of responsibility” (p. 30) Consistent with Peters (1998), consideration should be given to metacognitive and motivational factors if SDL is to be a viable and relevant concept to education in 21st century. It would seem that, for any student, SDL is predicated on the student having earned how to learn and having acquired the necessary epistemic and metacognitive awareness. If SDL is about controlling one’s learning and achieving meaningful and worthwhile outcomes, then epistemic cognition (i.e., cognitive development), metacognitive awareness, and motivational factors are preconditions.

Garrison’s (1997) model of SDL has considerable potentials to integrate issues related to motivational disposition (entering task), strategic planning (self-management), and metacognitive awareness (self-monitoring). It is strongly suggested that consideration be given to these higher-order cognitive dispositions and perspectives so that SDL can truly reflect autonomous learning. For SDL to have real pragmatic value in an educational context, the focus must shift to cognitive responsibility issues and transactional support for critical reflection. The ability of the teacher to provide control of the external learning tasks through well-designed learning packages must be balanced against concerns associated with modeling and diagnosing the internal cognitive (i.e., critically reflective) processes leading to higher-order outcomes (learning to learn).

ABD-EL-FATTAH594

Finally, the analyses of the present study indicated a relationship between SDL aptitude and academic achievement. Specifically, self-management was marked as the strongest predictor of academic achievement over Semester 1 and Semester 2, followed by self-monitoring. However, motivation significantly predicted academic achievement over the second semester only. These findings are consistent with what Stewart (2007) has recently reported that self-management was the strongest predictor of overall learning outcome ratings in project-based learning in engineering and that self-control and desire for learning had a positive relationship with overall learning outcome ratings but their R2 values were relatively low.

These findings can also be explained within the framework of the self-regulated motivational literature which indicates that successful learners have more effective and efficient learning strategies for accessing and using their knowledge, are self-motivated, and can monitor and change their strategies to improve their learning outcomes (Corno, 1989; Pintrich & DeGroot, 1990). When learners recognize their learning needs, formulate learning objectives, select contents, draw up learning strategies, procure teaching materials and media, identify additional human and physical resources and make use of them, and they themselves organize, control, inspect, and evaluate their own learning, they are more likely to perform highly on learning tasks.

Table 4Summary of results of the mediation analyses based on Baron and Kenny technique, Sobel test, and bootstrapping

Technique Statistics

Baron and Kenny

b SE TTotal effect path c Self-management (IV) > Self-monitoring (DV) .78 .07 11.14**Path aSelf-management (IV) > Motivation (M) .67 .08 8.38**Path bMotivation (M) > Self-monitoring (DV) .74 .07 10.57**Direct effect path c'Self-management (IV) > Self-monitoring (DV) .66 .08 8.25**Partial effect Gender (COV) > Self-monitoring (DV) .13 .11 1.18Academic major (COV) > Self-monitoring (DV) .17 .10 1.70Sobel Test

Path abIndirect effects of IV on DV through proposed mediators

Effect SE ZTotal .55 .10 5.5*Motivation .55 .10 5.5*Bootstrapping

Path abIndirect effects of IV on DV through proposed mediators

Data Boot Bias SETotal .65 .63 .02 .11Motivation .65 .63 .02 .11Bias Corrected and Accelerated Confidence Intervals

Lower UpperTotal .26 .73Motivation .26 .73

Note. N = 119. ‘Data’ is the indirect effect calculated in the original sample, ‘Boot’ is the mean of the indirect effect estimates calculated across all bootstrap samples. ‘Bias’ is the difference between “Data” and “Boot”. “SE” is the standard deviation of the bootstrap estimates of the indirect effect. This standard deviation could be used as a bootstrap-derived estimate of the standard error of the indirect effect. *p < .05, **p < .01.

GARRISON’S MODEL OF SELF-DIRECTED LEARNING 595

Zimmerman, Bonner, and Kovach (1996) suggested that one of the major advantages of using the self-management process is that it can improve not only one’s learning, but it can enhance one’s perception of self-confidence and control over the learning process. When one practices self-observing of his or her current learning, one can determine about the effective learning strategies to use. This process can help one achieve at higher levels and become a more self-directed and self-regulated learner.

In conclusion, self-direction is seen as essential if students are to achieve the ultimate educational goal of becoming continuous learners. Learning interests and opportunities for control can promote self-direction and continued learning. Opportunities for self-directed learning, in turn, can enhance metacognitive awareness and facilitate the conditions where students can learn how to learn.

References

Abd-El-Fattah, S. M. (2005). The effect of prior experience with computers, statistical self-efficacy, and computer anxiety on students’ achievement in an introductory statistics course: A partial least squares path analysis. International Education Journal, 5, 71-79.

Anderson, M. R. (1993). Success in distance education courses versus traditional classroom education courses. (Unpublished doctoral thesis). Oregon State University.

Bagozzi, R. P., & Youjae, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16, 74-94.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Boman, P., Smith, D. C., & Curtis, D. (2003). Effects of pessimism and explanatory style on development of anger in children. School Psychology International, 24, 80-94.

Brockett, R. G., & Hiemstra, R. (1991). Self-direction in adult learning: Perspective on theory, research, and practice. New York: Routledge.

Brookfield, S. (1985). Self-directed learning: A critical review of research. In S. Brookfiled (Ed.), Self-directed learning from theory to practice (pp. 5-16). San Francisco, CA: Jossey-Bass.

Brookfield, S. D. (1986). Understanding and facilitating adult learning. San Francisco, CA: Jossey-Bass.

Butler, D. L., & Winne, P. H. (1995). Feedback and self-regulated learning: A theoretical synthesis. Review of Educational Research, 65, 245-281.

Candy, P., Harri-Augstein, S., & Thomas, L. (1985). Reflection and the self-organized learner: A model of learning organizations. In D. Boud, R. Keough, & D. Walker (Eds.), Reflection: Turning experience into learning (pp. 100-116). London: Kogan Page.

Chernick, M. R. (1999). Bootstrap methods: A practitioner’s guide. New York: Wiley series in probability and statistics.

Chin, W. W. (1998). The partial least squares approach for structural equation modeling. In G. A. Macoulides (Ed.), Modern methods for business research (pp. 295-336). Mahwah, NJ: Lawrence Erlbaum.

Chin, W. W., & Newsted, P. R. (1999). Structural equation modeling analysis with small samples using partial least squares. In R. H. Hoyle (Ed.), Statistical strategies for small sample research (pp. 307-341). Thousand Oaks: CA: Sage Publications.

Corno, L. (1989). Self-regulated learning: A volitional analysis. In B. J. Zimmerman, & D. H. Schunk (Eds.), Self-regulated learning and academic achievement: Theory, research, and practice (pp. 111-141). New York: Springer-Verlag.

Crocker, L., & Algina, J. (1986). Introduction to classical and modern test theory. Forth Worth, TX: Harcourt Brace Jovanovich College Publishers.

Darmayanti, T. (1994). Readiness for self-directed learning and achievement of the students of Universitas Terbuka (Indonesian Open University). (Unpublished master´s thesis). University of Victoria, British Columbia, Canadá.

Field, L. (1989). An investigation into the structure, validity, and reliability of Guglielmino’s Self-Directed Learning Readiness Scale. Adult Education Quarterly, 39, 125-139.

Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39-5.

Garrison, D. R. (1991). Critical thinking and adult education: A conceptual model for developing critical thinking in adult learners. International Journal of Lifelong Education, 10, 287-303.

Garrison, D. R. (1993). An analysis of the control construct in self-directed learning. In H. B. Long (Ed.), Emerging perspectives of self-directed learning (pp. 27-44): Oklahoma: Research Center for Continuing Professional and Higher Education of the University of Oklahoma.

Garrison, D. R. (1997). Self-directed learning: Toward a comprehensive model. Adult Education Quarterly, 48, 18-33.

Gefen, D., Straub, D., & Boudreau, M. (2000). Structural equation modeling and regression: Guidelines for research practice. Communications of the Association for Information Systems, 7, 1-78.

Guglielmino, L. M. (1977). Development of the Self-Directed Learning Readiness Scale. (Unpublished doctoral thesis). University of Georgia.

Guthrie, J. T., McGough, K., Bennett, L., & Rice, M. E. (1996). Concept-oriented reading instruction: An integrated curriculum to develop motivations and strategies for reading. In L. Baker, P. Afflerbach, & D. Reinking (Eds.), Developing engaged readers in school and home communities (pp. 165-190). Hillsdale, NJ: Erlbaum.

Haggerty, D. L. (2000). Engaging adult learners in self-directed learning and its impact on learning styles. (Unpublished doctoral thesis). University of New Orleans.

Haenlein, M., & Kaplan, A. M. (2004). A beginner’s guide to partial least squares analysis. Understanding Statistics, 3, 283–297.

ABD-EL-FATTAH596

Haron, S. (2003). The Relationship between readiness and facilitation of self-directed learning and academic achievement: A comparative study of web-based distance learning models of two universities. (Unpublished doctoral thesis). University of Putra, Malaysia.

Harriman, J. K. (1990). The relationship between self-directed learning readiness, completion, and achievement in a community college telecourse program. (Unpublished doctoral thesis). University of Georgia.

Hsu, Y. C., & Shiue, Y. M. (2005). The effect of self-directed learning readiness on achievement comparing face-to-face and two-way distance learning instruction. International Journal of Instructional Media, 32, 143-155.

Hulland, J. (1999). Use of partial least squares (PLS) in strategic management research: A review of four recent studies. Strategic Management Journal, 20, 195-204.

Kanfer, R. (1989). Cognitive processes, dispositions, and behavior: Connecting the dots within and across paradigms. In R. Kanfer, P. L. Ackerman, & R. Cudeck (Eds.), Abilities, motivation, and methodology (pp. 375-388). Hillsdale, NJ: Lawrence Erlbaum.

Kline, P. (1986). A Handbook of test construction: Introduction to psychometric design. London: Methuen.

Knowles, M. (1975). Self-directed learning. New York: Association Press.

Long, H. B. (1989). Self-directed learning: Merging theory and practice. In H. B. Long (Ed.), Self-directed learning: Merging theory and practice (pp. 1-12). Oklahoma: Research Center for Continuing Professional and Higher Education of the University of Oklahoma.

Morris, L. S. (1995). The relationship between self-directed learning readiness and academic performance in a nontraditional higher education program. (Unpublished doctoral thesis). University of Oklahoma.

Morrow, L. M., & Young, J. (1997). Parent, teacher, and child participation in a collaborative family literacy program: The effects of attitude, motivation, and literacy achievement. Journal of Educational Psychology, 89, 736-742.

Nunnally, J. C. (1978). Psychometric theory (2 ed.). New York: McGraw-Hill.

Ogazon, A. G. (1995). The contribution of self-directed learning readiness to the achievement of junior students at a branch of the state of Florida university system. (Unpublished doctoral thesis). Florida International University.

Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, and Computers, 36, 717-731.

Peters, O. (1998). Learning and teaching in distance education: Analyses and interpretations from an international perspective. London: Kogan Page.

Pintrich, P. R., & DoGroot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82, 33-4.

Prawat, R. S. (1992). Teachers’ beliefs about teaching and learning: A constructivist perspective. American Journal of Education, 100, 354-395.

Resnick, L. B. (1991). Shared cognition: Thinking as social practice. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 1-20). Washington, DC: American Psychological Association.

Ringle, C. M., Wende, S., & Will, A. (2005). SmartPLS 2.0 (M3) Beta. Hamburg, Germany: University of Hamburg.

Savoie, M. M. (1979). Continuing education for nurses: Predictors of success in courses requiring a degree of learner self-direction. (Unpublished doctoral thesis). University of Toronto, Canadá.

Sellin, N., & Keeves, J. (1997). Path analysis with latent variables. In J. P. Keeves (Ed.), Educational research, methodology, and measurement: An international handbook (pp. 633–640). Oxford, UK: Pergamon.

Sharicey, S. B., & Sharpies, A. Y. (2001). An approach to consensus building using the Delphi technique: Developing a learning resource in mental health. Nurse Education Today, 21, 398-34.

Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. In S. Leinhardt (Ed.), Sociological methodology (pp. 290-312). Washington DC: American Sociological Association.

Stewart, R. A. (2007). Investigating the link between self-directed learning readiness and project-based learning outcome: The case of international masters students in an engineering management course. European Journal of Engineering Education, 32, 453-465.

Straka, G. A. (1995). Problems of measuring self-directed learning readiness. Paper presented at the Asia-Pacific Seminar on Self-directed Learning, Seoul, Korea.

Temple, C., & Rodero, M. L. (1995). Active learning in a democratic classroom: The “Pedagogical invariants” of Celesting Freinet (Reading around the world), Reading Teacher, 49, 164-167.

Thompson, D. (1992). Beyond motivation: Nurses’ participation and persistence in baccalaureate nursing programs. Adult Education Quarterly, 42, 94-105.

Velicer, W. F., & Jackson, D. N. (1990). Component analysis versus common factor analysis: Some issues in selecting an appropriate procedure. Multivariate Behavioral Research, 25, 1-28.

Weisz, J. R. (1983). Can I control it? The pursuit of veridical answers across the life-span. In P. B. Baltes, & O. G. Brim (Eds.), Life-span development and behavior (pp. 233-300). New York: Academic Press.

Zimmerman, B. J., Bonner, S., & Kovach, R. (1996). Developing self-regulated learners: Beyond achievement to self-efficacy. Washington, DC: American Psychological Association.

Received June 2, 2009Revision received November 11, 2009

Accepted December 18, 2009