Foundations for Personalized Web Learning Environments

Margaret Martinez, Ph.D.
Tucson, AZ 85737

C.Victor Bunderson
Edumetrics Institute and Brigham Young University
Provo, UT 84602

KEYWORDS
Learning Differences, Conative, Affective, E-Learning, WBT, Learning Theory

I. INTRODUCTION

The Web offers the perfect technology and environment for personalized learning where learners can be uniquely identified, content can be specifically presented, and progress can be individually monitored, supported, and assessed.  Technologically, researchers are making rapid progress towards personalized learning on the Web using object architecture and adaptive technology. However, missing still is a whole-person understanding of how individuals learn online (more than just how they process, build, and store knowledge).  Primarily cognitive solutions originally designed for the classroom solutions (and facilitated by instructors) are often not enough to meet the individual, sophisticated needs of Web learners.

Offering an alternative perspective about learning on the Web, this paper describes a research foundation that supports individual differences from a more personal level.  It discusses (a) sources for individual learning differences, (b) specific reasons why some learners may be more self-directed or self-motivated than others, and (c) design guidelines that tap into the dominant influence of emotions, intentions, and social aspects on learning.  These insights offer simple ways to enhance and evaluate contemporary Web instructional designs so that they support personalized needs and instill the right habits for improved learning and performance.

This paper is aimed at readers seeking new perspectives for understanding individual differences and personalized learning on the Web.  The purpose is to suggest that after years of research focused on primarily cognitive models, we have learned that these solutions have often proved unpredictable and unstable, especially for online learning.  Reeves [1] advocated stronger, more reliable theoretical foundations when he suggested that "much of the research in the field of computer-based instruction is pseudoscience because it fails to live up to the theoretical, definitional, methodological, and/or analytic demands of the paradigm upon which it is based."

In contrast, conative (desires, intentions) and affective (emotions, feelings) attributes of persons are more stable over different online learning situations.  Consequently, many Web learning designers are finding that conventional cognitive solutions are not enough.  They are discovering the need to increase their focus on the conative and affective factors that influence learning.  In this context, the purpose of this paper is to examine higher-order psychological influences on learning.  This perspective leads to an examination of the dominant impact of emotions and intentions on cognitive processing.  The paper considers (a) vital relationships between key psychological factors (conative, affective, cognitive, and social) which influence learning differently, (b) critical links between Web learning environments, learning differences and learning ability, and (c) supportive Web learning environments that match individual learning differences.

II. BACKGROUND

We are still very much in the experimental stage for creating Web learning environments.  The completion rates for Web learning are notoriously low.  More needs to be learned about designing successful online environments, technically, pedagogically and personally.  In the fifties, Cronbach [2] challenged the field to find "for each individual the treatment to which he can most easily adapt." He suggested that consideration of the treatments and individual together would determine the best payoff because we "can expect some attributes of person to have strong interactions with treatment variables. These attributes have far greater practical importance than the attributes which have little or no interaction."

In dividing pupils between college preparatory and non-college studies, for example, a general intelligence test is probably the wrong thing to use. This test being general, predicts success in all subjects, therefore tends to have little interaction with treatment, and if so is not the best guide to differential treatment. We require a measure of aptitude that remains to be discovered. Ultimately we should design treatments, not to fit the average person, but to fit groups of students with particular aptitude patterns.  Conversely, we should seek out the aptitudes which correspond to (interact with) modifiable aspects of the treatment.

A research program on the interactions between cognitive abilities and learning was conducted by Bunderson and Dunham [3] during this same decade.  Like the results found by Cronbach [4] and Cronbach and Snow [5], interactions of instructional treatments with cognitive aptitudes were shown to be inconsistent and hard to replicate.  Moreover, these investigators observed that excellent instructional design strategies could reduce or wholly remove the impact of cognitive abilities on learning.  Using instructional design methods combined with laboratory and lesson-like studies of learning complex concepts, they stated:  "We have learned how to reduce or to remove the constraint on learning associated with low scores on certain aptitude tests" [3 (p.34)].  These lessons about the power of good instructional design over aptitude x treatment interactions were taken over into the design of the TICCIT system, an early computer-based instructional system.  Instead of the all-knowing psychologist prescribing the best treatment based on an aptitude pattern, learners were given choice and control over the instructional sequence and could select their own tactics through a learner control language [6], [7].   It was found in a decade of studies using the TICCIT system that some learners disliked the extensive learner control provided, and some loved it.  At that time of cognitivist thinking the notion of a whole-person learning orientation construct that would enable prediction of which students would thrive in a learner control environment and which would not had not occurred to the TICCIT researchers.  However, the TICCIT designers opened the doors toward conation and affect in the design goals of TICCIT [8].  In addition to the goals of mastery, efficiency and improved learning strategies, TICCIT's designers sought Approach rather than Avoidance, a purely conative outcome, and Responsibility toward scheduling their own learning and exerting continual effort, a conative/affective goal that is now an important part of the Learning Orientation Construct.


A. Research Between 1960 and 1970
In 1965, Gagné organized a major conference to discuss and explore individual differences in learning [9 (p. xi)].   During the conference, Melton [9 (p. 239)] suggested "that we frame our hypotheses about individual difference variables in terms of the process constructs of contemporary theories of learning and performance."  In retrospect, the conference's most important consensus was that conceptual formulations of processes or mechanisms, that is, information or knowledge processing, intervened between stimuli and response-the prevalent behavioral learning perspective [10 (p. 3)]. Critical to this consensus was the view that intelligence and achievement relied heavily on specific intrinsic cognitive processing. "This conference reflected a change in the conceptualization of intelligence as measured performance to mental mechanisms" [10 (p. 3)].

Between 1960 and 1970, Cronbach [4], [5] and others "searched fruitlessly for interactions of abilities." They were looking for "aptitudes" (characteristics that affect responses to the treatment) that explained how to instruct students one way and not another (i.e., evidence that showed regression slopes that differed from treatment to treatment).  In the seventies, Cronbach [4] still advocated that a closer scrutiny of cognitive processes would be a profitable next phase of work on Aptitude Treatment Interactions (ATIs). He highlighted research that related success to the Ai (Achievement via Independence) and Ac (Achievement via Conformance) scores of Gough's California Personality Inventory. The evidence continued to show that the learning outcomes were better when the instructor's presentation adapted to the student's aptitude and personality [5]. For example, the "constructively motivated student who seeks challenges and takes responsibility is at his best when an instructor challenges him and then leaves him to pursue his own thoughts and projects."  Cronbach [4] continued to emphasize the important relationship between cognitive aptitudes and treatment interactions.  Nevertheless, he states that "Snow and I have been thwarted by the inconsistent findings coming from roughly similar inquiries. Successive studies employing the same treatment variable find different outcome-on-aptitude slopes." He surmised that the inconsistency came from unidentified interactions. Finally, Snow and Cronbach [5] concluded that "an understanding of cognitive abilities considered alone would not be sufficient" to explain learning, individual learning differences, and aptitude treatment interactions.

B. Research in the 1980's
In the early eighties, the cognitive process analysis of aptitudes processes continued with variations. Snow [11] described the ATI investigation as process-oriented research on individual differences in learning and cognition. Although he and Cronbach were looking for a "whole-person view" of learning, Snow believed that it was primarily the cognitive processes that should be considered in the design and development of adaptive instructional systems. Eventually the new "aptitudes" evolved into cognitive styles (today, often called learning styles) to represent the predominant modes of information processing (i.e., preferred learning sets to the acquisition, retention and retrieval of new knowledge).

In the late eighties, Snow [12] described how in cognitive psychology conation as a learning factor has been "demoted" and "since it seems not really to be a separable function," it is merged with affection. Together these factors are viewed as "mere associates or products of cognition" and then ignored. He warned that individual difference constructs or "aptitude complexes" needed greater consideration of the joint functioning between cognitive, conative, and affective processes. Snow was in search of an information-processing model of cognition that would include possible cognitive-conative-affective intersections. He was looking for a way to fit realistic "aspects of mental life, such as mood, emotion, impulse, desire, volition, purposive striving" into instructional models. According to Snow [13], the best instruction involves treatments that differ in structure and completeness and high or low general ability measures. Highly structured treatments (e.g., high external control, explicit sequences and components) seem to help students with low ability but hinder those with high abilities (relative to low structure treatments).

C. Current Research Activities
Cronbach's and Snow's research, like the multidisciplinary research of many others, set the stage for the learning orientation research.  For example, recent neuroscience research is revealing the amygdala/hippocampus' (the brain's emotional system) important influence on learning and memory [14], [15]. The learning orientation research attempts to reveal the dominant power of emotions and intentions on guiding and managing cognitive processes (no longer demoted to a secondary role). It is in understanding the structure and nature of the complex relationships between learning orientations and interactions that we can return to Cronbach's original hypothesis that we should find "for each individual the treatment to which he can most easily adapt" [2]. And, ultimately we should design treatments, not to fit the average person, but to fit groups of students with particular aptitude patterns.  This is a personalization or adaptive learning approach (called mass customization) that identifies aggregate types or segmented populations.  Conversely, we should seek out the aptitudes which correspond to (interact with) modifiable aspects of the treatment [2].

As can be expected the new lines of research will continue to reopen the old questions, gain from the research accomplished in the past, and pose exciting new questions for the future. As we look forward to new issues highlighting the importance of emotional and intentional states on cognitive processing, waiting in the wings to be discovered are the treatments that lead toward more successful learning and performance. And perhaps, as some may have already predicted, the hegemony of cognition over intent and affect is coming to an end.

Meanwhile, many contemporary researchers have extended their research on learning and memory constructs (and associated measures) to include conative, affective and social influences [16], [17], [18], [19], [20], [21].  Still, most have done so without recognizing and incorporating the dominant influence of conative, affective, and social factors. As a result, powerful psychological factors, such as intentions, personal desire, will, striving, motivation, efficacy, collaboration, pride, fear, frustration and satisfaction, are still being ignored or demoted to a secondary role.  The cognitive-rich tradition remains the dominant consideration for learning.

III. PERSONALIZATION

Today's researchers and designers alike are seeking more sophisticated learning theories based on proven research showing how the brain works.  Recent developments in neuroscience are revolutionizing our understanding of how individuals really learn.  These more sophisticated theories and learning and memory constructs are leading the way for personalizing or adapting online learning environments and instruction.  A key consideration in personalization or adaptive learning is determining dominant or higher-level sources for individual learning differences.  This involves understanding how the brain's emotional system influences cognitive processes or how we think and learn. Much of our evolving understanding and research on individual learning differences remains broadly focused on cognitive interests and intrinsic or extrinsic mechanisms for information processing and knowledge building.  Hence, consideration of an important piece of learning is missing, since primarily cognitive solutions often overlook fundamental whole-person learning needs (such as the dominant influence of emotions and intentions) for self-directed and self-motivated learning.  The cognitive solutions generally support traditional roles where an instructor manages emotions, intentions and social issues for the common majority, as learners pursue cognitive solutions.

Traditional classroom (primarily cognitive) solutions, although often used, are not always viable online solutions.  Online, learners need to want and intend to become more self-supporting and self-directing learners, independent of the instructor.  Too many students emerging from classroom environments are ill-equipped to handle online learning environments.  Recognizing the online learning ability gap and providing solutions that consider the whole-person perspective is a step in helping student transition to more successful, self-directed online learning. It is not surprising that completion rates are low since the majority of today's learners are conditioned to rely on instructors.  Schools and industry require a more sophisticated understanding of the psychological characteristics of learning to change this conditioning.  Especially important is learning the helps learners want and intend to improve performance and negotiate constant improvement and change, independently, passionately and productively.  More personalized learning is a step in this direction.  As companies decide on next-generation e-learning alternatives, they need to first understand the dominant power of emotions and intentions on learning, and second, seek personalized solutions that use this understanding to revolutionize the presentation of learning and performance solutions.

IV. LEARNING ORIENTATION THEORY

This paper uses Learning Orientations  [22] to describe dominant sources  (i.e., emotions and intentions along with cognitive and social factors) for learning differences. Learning orientations developed and examined during previous research [23], represent how individuals (aggregated by varying beliefs, emotions, intentions and ability), plan and set goals, commit and expend effort, and then experience learning to attain goals.

This is an attempt to capture aspects of human learning that go beyond conventional constructs of cognitive ability.  The Learning Orientation Theory [24] hypothesizes that understanding the depth of an individual's fundamental emotions and intentions about why, when and how to use learning and how it can accomplish personal goals or change events is fundamental to understanding how successfully the individual learns, interacts with an environment, commits to learning, performs, and experiences learning and change.  In contrast, how well instructors and course designers understand and match learning orientations, is, in turn, how well they can present instruction that fosters self-motivation, encourages online relationships, and supports successful learning and performance.

A. Learning Orientation Model
The Learning Orientation Theory provides guidelines for developing learner-difference profiles, called learning orientations, which describe fundamental individual learning differences.  Orientations generally represent an individual's approach to learning-to differing degrees of success. 

The Learning Orientation Model [22] highlights the whole-person perspective as it presents ranges for four learning.

• Transforming Learners
• Performing Learners
• Conforming Learners
• Resistant Learners

Based on published research [22], [23], [24], [25], [26], [27], [28], the four learning orientations (see Table 1) distinguish learning variability and describe how individuals follow a complex mix of beliefs, desires, emotions, intentional effort, and cognitive and social styles to learn.  Learning orientations are how individuals, with varying beliefs, values, and levels of ability, intentionally approach, commit and apply effort, and then experience learning to attain short- or long-term goals. They describe the individual's proclivity to take control, set goals, attain standards, manage resources, solve problems and take risks to learn. 

Learners situationally fall along the continuum of learning orientations.  They may move downwards or upwards (vertically and horizontally) in response to negative or positive responses, conditions, resources, results and experiences.  For example, upward movement into higher orientations requires far greater effort, learning autonomy, intentions, feelings and beliefs about learning than downward range movement.  Learning orientations are an effective way to differentiate the audience according to the higher-order psychological factors that powerfully impact learning and performance.  They describe how we (influenced by emotions or intentions) foster, develop, manage and sometimes override our cognitive learning preferences, strategies and skills.  Learning orientations (a) represent conative, affective, cognitive and social influences on learning from a whole-person perspective, (b) introduce higher-order psychological aspects into audience analysis and instructional design methodology, (c) provide guidelines for differentiating the audience, and (d) help designers tailor solutions that improve learning ability and the learning experience.

The profiles for learning orientations in Table 1 use the three construct factors to describe how learners, following beliefs, values, emotions and intentions self-motivate themselves to learn (1. Conative/Affective factor), contribute efforts (2. Strategic Planning and Committed Effort factor), and self manage learning (3. Learning Autonomy factor) to varying degrees. 

Table 1.  Descriptions for Four Learning Orientations.

B. Design Guidelines for Personalized Learning
Instructional design for Web learning should address the unique sources for learning differences from a whole-person perspective.  In some ways, designs should emulate the instructor's experienced, intuitive ability to recognize and respond to how individuals learn differently.  Certainly designs should foster interest, value, and encourage more self-motivated, self-directed learning.  Matching a more personalized solution to individual differences, identified through audience analysis, should become an integral part of the entire instructional design process. How to also introduce and support conative, affective, and social factors in instruction is the challenge.  Table 2 suggests possible guidelines using each of three learning orientations.  These suggestions are helpful in planning instruction, promoting interactivity, capturing interests, designing interfaces and environments, delivering instruction, practice, feedback, and assessment, helping learners monitor progress, evaluating performance, and making revisions.

Table 2. Instructional Strategies for Three Learning Orientations.

V. CONCLUSIONS

Hopefully, these suggestions will contribute to more successful learning via the Web and a greater understanding about fundamental learning differences influenced by conative and affective influences (how the brain works).  When we design a course with only a universal type of learner in mind (all with similar emotions and intentions) we unintentionally set learners up for frustration and possible failure.  If we are serious about providing good online instruction for learners, we must provide multiple ways to provide instruction and environments so that all learners will want to learn on the Web and continue to have opportunities for success.  The benefits of personalizing learning to individual differences particularly address important human issues previously managed by instructors in the classroom (for an example, an instructor that can see frustration, lack of confidence, mistakes, impatience, reactions, and boredom).

The descriptions in Table 1 and 2 are a step in recognizing and accommodating individual learning differences.  They offer designers a blueprint based on research foundations and provide specific targets and measures to monitor performance and predict and foster more successful learning outcomes.   These descriptions are also an important step in recognizing the expanded, dominant role and impact of emotions and intentions on learning, especially as we help learners become more sophisticated, self-motivated, and self-directed learners. 

REFERENCES

1. Reeves, T., Pseudoscience In Computer-Based Instruction: The Case of Learner Control Research, Journal of Computer-Based Instruction, Vol. 20, No. 2, pp. 39-46, 1993.
2. Cronbach, L., The Two Disciplines of Scientific Psychology, American Psychologist, pp. 671-684, 1957.
3. Bunderson, C.V., and Dunham, J.L.,  Research Program on Cognitive Abilities and Learning,  Final Report, Advanced Research Projects Agency Contract N00014?67?A?026?0006, Computer?Assisted Instruction Laboratory,  The University of Texas at Austin, 1970.
4. Cronbach, L., Beyond the Two Disciplines of Scientific Psychology, American Psychologist, 12, pp. 116-127, 1975.
5. Cronbach, L., and Snow, R., Aptitudes and Instructional Methods: A Handbook for Research on Interactions, New York: Irvington Publishers, 1977.
6. Merrill, M.D., Schneider, E. W., and Fletcher, K.A., TICCIT, Volume 40 in the Instructional Design Library, Educational Technology Publications, Englewood Cliffs, New Jersey, 1980.
7. Bunderson, C.V., The Design and Production of Learner-controlled Courseware for the TICCIT System.  International Journal of Man-Machine Studies, 6, pp. 479?491, 1974.
8. Bunderson, C. V., The TICCIT Project: Design Strategy for Educational Innovation.  In S. A. Harrison & L. M. Stolurow (Eds.), Productivity in Higher Education, National Institutes of Education, September 1973.
9. Melton, A., Individual Differences and Theoretical Process Variables: General Comments on the Conference. In R. M. Gagné (Ed.), Learning and Individual Differences, Columbus, Ohio, Merrill, 1967.
10. Federico, P., Adaptive Instruction: Trends and Issues. In R. Snow and M. Farr (Eds.), Conative and Affective Process Analyses, Vol. 1, pp. 1-26, Hillsdale, NJ, Lawrence Erlbaum Associates, 1980.
11. Snow, R.E., Aptitude Processes. In R. Snow, P.-A. Federico, and W. E. Montague (Eds.), Aptitude, Learning, and Instruction: Cognitive Process Analyses of Aptitude, Vol. 1, pp. 27-63, Hillsdale, NJ, Erlbaum, 1980. Available: http://www.aqps.qc.ca/bulletin/04/04-01-02.htm
12. Snow, R., and Farr, M., Cognitive-Conative-Affective Processes in Aptitude, Learning, and Instruction: An Introduction. In R. Snow and M. Farr (Eds.), Conative and Affective Process Analyses, Vol. 3, pp. 1-10, Hillsdale, NJ, Erlbaum Associates, 1987.
13. Snow, R., Aptitude-Treatment Interaction as a Framework on Individual Differences in Learning. In P. Ackermann, R.J. Sternberg, and R. Glaser (Eds.), Learning and Individual Differences, New York, W.H. Freeman, 1989.
14. Ledoux, J., Emotional Brain: The Mysterious Underpinnings of Emotional Life, New York, Simon and Schuster, 1996.
15. Cahill, L. and McGaugh, J.L., Mechanisms of Emotional Arousal and Lasting Declarative Memory, Trends in Neurosciences, Vol. 21, pp. 294-299, 1998.
16. McCombs. B., Alternative Perspectives For Motivation, In Developing Engaged Readers in School and Home Communities. In L. Baker, P. Afflerbach, and D. Reinking (Eds.), Hillsdale, Lawrence Erlbaum Associates, pp. 67?87, 1996.
17. Bereiter, C., and Scardamalia, M., Surpassing Ourselves: Inquiry into the Nature and Implications of Expertise, Chicago: Open Court, 1993.
18. Corno, L., and Mandinach, E. B., The Role of Cognitive Engagement in Classroom Learning and Motivation, Educational Psychologist, 18(2), pp. 88?108, 1983.
19. Deci, E., and Ryan, R., A Motivational Approach to Self: Integration in Personality, In R. Dienstbier (Ed.), Nebraska Symposium on Motivation: Vol. 38. Perspectives on motivation (pp. 237-288). Lincoln: University of Nebraska Press, 1991.
20. Wiener, R., Attribution Theory, Achievement Motivation, and the Educational Process. Review of Educational Research, Vol. 42, pp. 203-215, 1972.
21. Brown, A., Metacognition, Executive Control, Self-regulation, and Other More Mysterious Mechanisms, In F. Weinert and R. Kluwe (Eds.), Metacognition, Motivation, and Understanding, pp. 65-116, Hillsdale, NJ, Erlbaum Associates, 1987.
22. Martinez, M., Learning Orientations Model. Available: http://www.trainingplace.com/source/research/learningorientations.htm Accessed Dec. 9, 2000.
23. Martinez, M., An Investigation into Successful Learning: Measuring The Impact of Learning Orientation, A Primary Learner-Difference Variable, on Learning. Dissertation. (University Microfilms No. 992217) 1999.
24. Martinez, M., Successful Learning Research Community. Available: http://www.trainingplace.com/source/research/learningorientations.htm#lo Accessed Dec. 9, 2000.
25. Martinez, M., Key Design Considerations for Personalized Learning on the Web, Education Technology & Society, October 2000 (in press).
26. Martinez, M., and Bunderson, C. V., Building Interactive Web Learning Environments to Match and Support Individual Learning Differences, Journal of Interactive Learning Research, Vol. 11, No. 2, 2000.
27. Martinez, M., Research Design, Models, and Methodologies for Examining How Individuals Successfully Learn on the Web, Special Research in Technical Communication, Vol. 46, No. 4, pp. 470-487, 1999.
28. Martinez, M., Mass Customization: A Paradigm Shift for The 21st Century, ASTD Technical Training Magazine, Vol. 10, No. 4, pp. 24-26, 1999.
    
    

ABOUT THE AUTHORS

Margaret (Maggie) Martinez, CEO at The Training Place has worked in the fields of learning, information and technology for more than fifteen years.  Previously she was the Worldwide Training and Certification Director for WordPerfect Corporation.  Martinez has provided leadership, insight and perspective on learning issues to major corporations worldwide as they cope with rapidly changing business opportunities, performance improvement, and accelerated technological advancement.  Ms. Martinez' professional initiatives have focused on demystifying the world of learning and performance by pioneering individual learning difference and personalization research.  This research explores the powerful impact of emotions and intentions on learning and performance.  She has a Ph. D. in Instructional Psychology and Technology, regularly presents at major conferences, and publishes in academic and trade publications. 

C. Victor Bunderson,  Ph.D. Princeton, is Professor of Instructional Psychology & Technology at Brigham Young University and Chairman of Alpine Media and the Edumetrics Institute.  His interests include research on computer-administered measurement and assessment, integration of continuous progress measurement with computer-aided systems for instruction and learning, Learner Control / Intentional Learning and e-learning models, and developing expert learning environments -- defining new roles, and using new tools to facilitate technology. Dr. Bunderson served as Vice President for Research Management at Educational Testing Service where he directed R&D leading to a new generation of systems that integrate new forms of assessment with instruction.  Current research investigates measurement of learning progress over time on invariant scales of learning and growth.