Developing the Metacognitive and Problem-Solving Skills of ...

Chapter4

Developing the Metacognitiveand Problem-Solving Skills

of Science Students in Higher Education

Rowan W. HollingworthCatherine McLoughlin

INTRODUCTION

In tertiai-y education nowadays a greater emphasis is being di-rected toward the development of generíc skills or graduate attributes,including communication skills, global perspectives, problem solv-ing, teamwork, and social responsibility. Reliance on a content-basedcurrículum in science is not an appropriate preparation for the rapidlychanging world of the future (Lowe, 1999). The new emphasis on ge-neric skills is aimed at addressing an urgent need for professionalswho can find realistic solutions to complex, real-world problems.Jonassen (2002) goes so far as to .state, "1 believe that the only 1egiti-mate goal of professional education, either in universities or profes-sional training, is problem solving" (p. 78). It is clear, then, that ter-tiary educators need to carefully examine their methods of teachingproblem-solving, as well as the types of problems they select for theirstudents, if they wish to produce graduales effective in the modernworkplace, society, and life in general.

In this chapter the development of problem-solving skills, specifi-cally in the context of first-year university-Ievelscience subjects, isdiscussed. This is done in the context of the broader profile of stu-dents now entering universities and studying in both fue on-campusand off-campus (distance-education) moJes. The types of problemswhich students may tackle in their learning activities are considered.The teaching/learning environments in which learning tasks are car-

63

64 TEACHINGIN THESCIENCES

ried out and which will effectively support the development of goodproblem-solving sldlls are examined. Finally, a number of approachesto developing the higher-order cognitive sldlls of first-year sciencestudents, including an online tutorial, metAHEAD, are described.

LINKING PROBLEM-SOL VINGAND METACOGNITIVE SKILLS

Broadly spealdng, the development of effective problem-solvingsldlls depends on two factors:

. The teachinglleaming environment in which the problem-solvingsldlls are developed. The types of problem students are exposed to

In terms of pedagogy, it is not sufficient to rely on a "transmission ofknowledge" approach, which relies heavily on regurgitation of factsand concepts and the solution of routine problems or exercises. Thistype of approach targets only lower-order cognitive skills (Zoller,2000). Even so, the traditional "show and tell, then practice" ap-proach to problem solving may help students develop some fluency athandling and applying concepts. However, tbis simplistic approachcan no longer be relied on as the sale strategy or even a very effectivestrategy for building tbe bigber-order cognitive skills required totackle tbe more ill-defined, complex, interdisciplinary problems uni-versity graduales now face when they enter the workforce.

Recent analysis of researcb on the teaching ofproblem solving hassbown tbat knowledge of strategy and practice of problem solving hasbad little effect on student perfonnance and acbievement, wbereas ef-fective approaches to teaching problem solving alJ gave attention tocontextuaJized strategies related to the knowledge base. The leamingconditions recognized as significant for building problem-soIvingsldIIs are tbose that provide Iearners witb guideIines and criteria theycan use in judging tbeir own problem-solving processes and producíS(Everson and Tobias, 1998). A repertoire of Ieaming strategies, a ca-pacity to manage one's own leaming, and an awareness of one's ownknowIedge and skills are fundamental in arder to leam effectively andto problem salve in a variety of contexts. The range of skills relating

,

Developing Metacognitive ami Problem-Solving Skills 65

to the self-management ofIeaming as students engage in monitoringand evaluating their own problem solving is known as metacognition.

Development of Metacognition

Metacognition invoIves both knowledge of cognition (the leamers'knowledge abOlirtbeir own processes of cognition) and regulation ofcognition (tbe ability to monitor and control tbose processes) (Met-calfe and Shimamura, 1994; Schraw, 1998). Metacognition can beviewed as a supervisory or metalevel system that controls and re-ceives feedback from nonnal infonnation processing. Metacognitiveknowledge refers to wbat the learner knows and understands abolirthe task in baTId,whiIe metacognitive regulation refers to the strate-gies the Iearner uses to complete the task. Tbis regularían involvesplanning, organizing,. and monitoring tbe task, but it also involvesevaluating outcomes and reflecting on leaming and problem sol ving.

Tbe literature attests to the fact tbat even at beginning tertiary leve1,few students arrear to have developed the expert problem-solvingand metacognitive sldIIs that enable them to cope effectively withlearning independently and successfuIIy in tbe sciences (Volet,McGill,and Pears, 1995; Everson and Tobias, 1998; Gourgey, 1998). There isevidence that metacognitive sldlls can be taught, althougb the rangeof programs and approaches attempted has been varied. Gredler(1997) propases three essential conditions, which appIy for trainingand development of metacognitive sldIIs:

. Tbe training sbould involve students' awareness ofwbat tbe pro-cess involves, as this makes them participants in the process.. The performance criteria used for evaluation of achievementsbould match tbe ldnds of metacognitive activities addressed intbe instruction.

. Metacognitive training should provide support for engagementin metacognitive activities.

Masui and De Corte (1999) propase similar conditions, suggestingan integrated ser of instructionaI principIes for an effective leamingenvironment to enhance metacognition and problem-solving skillsfor university students:

66 TEACHING IN THE SCIENCESDeveloping Metacognitive and Problem-Solving Skills 67

TEACHING PROBLEM SOLVING

rather than the field of cognitive psychology, in which bullan prob-lem solving is a well-established and wide research arca. Much of fueresearch on the training of problem solving and metacognition has in-volved primary- and secondary-Ievellearners, rather than those at thetertiary level.

Sadly, it appears that this large research resource has been onlylightly tapped by the majority of university science teachers. Gabel(1994) provides thorough reviews of research on problem solving inseveral science subjects, which had been carried out up to the mid-1990s. More recently Taconis, Ferguson-Hessler, and Broekkamp(2001) analyzed articles researching the effectiveness of teachingstrategies for science problem solving as reported in twenty-fivehigh-quality intemationaljoumals between 1985 and 1995. From ap-proximately 2,000 papers in these joumals, forty experiments intwenty-two papers on the teaching of prob1em solving were used.These studies concentrated on the cognitive aspects of the teachinginterventions and left the metacognitive aspects implicit.

The traditional approach to problem solving in science has been toengage students in repetition of routíne exercises, with an emphasison the use of algorithms and sequences of steps rather than strategiesand reflection on processes. Hobden (1998) suggests this has beenused uncritically as a teaching strategy "on the optimistic assumptionthat success with numerical problems breeds an implicit conceptualunderstanding of science" (p. 219). This approach may help stuqentsdevelop routine expertise, that is, speed and accuracy at routine prob-lem solving, but will fail to develop adaptive expertíse, the ability toreflect on strategies or to adapt to solving new problems in a flexiblemanner (Hatano and Inagaki, 1986).

From their analysis, Taconis, Ferguson-Hessler, and Broekkarnp(2001) found that attention to the structure and function ofthe knowl-edge base was a rearme of effective metacognitive training, while at-tention to knowledge of strategy and the practice of problem solvingappeared to have little effect. The 1eaming conditions recognized assignificant for building prob1em-solving skills are those which pro-vide leamers with guidelines and criteria they can use injudging theirown problem-solving processes and prodUCíS.The provision of im-mediate feedback to leamers is also essential. These conc1usions are

congruent with earlier research carried out by researchers in the field

. Embed acquisition of knowledge and skills in a real study con-text.. Take joto account the study orientarían of students and theirneed to experience the relevance of the learning and study tasksoffered to them. ... Sequence teaching methods and leaming tasks and interrelatethem.. Use a variety of forms of organization or social interaction.. Take joto account informal prior knowledge and individual dif-ferences between students.. Learning and thinking processes should be verbalized and re-flected upon.

Examples of studies of metacognitive development at the univer-sity level are rather scarce. They typically have involved lecturers andinstructors in long-term, face-to-face situations ayer a period of atleast one semester. These inc1ude a study of first-year computer sci-ence students' development of a metacognitive strategy and coachingits use in a socially supportive environment (Volet, 1991), and a self-directed 1eaming program to develop transferable 1earning and meta-cognitive skills for first-year chemistry students (Zeegers, Martin,and Martín, 1998). Outside the science arca, Masui and De Corte(1999) have examined the trainability and effect on academic perfor-mance of enhancing the 1earning and problem-solving skills of busi-ness economics students.

There is consensos in the research literature that 1eamers have an

opportunity to evaluate the outcome of their efforts, to reflect on andself-assess their own approaches to 1earning. Simply providing knowl~edge without experience or vire versa does not seem to be sufficientfor the development of metacognitive control. The most effectivemetacognitive instruction schemes in the literature involve providingthe learner with knowledge of cognitive processes and strategies, to-

gether with experience or practice in using them (Boekaerts, Pintrich,and Zeidner, 2000).

The focos in this chapter is directed more to the field of science ed-ucation and the rapidly developing field of educational technology,

68 TEACHINGIN THESCIENCES Developing Metacognitive ami Problem-Solving Skills

TABLE 4.1. Some Characteristics of Different Types of Problems

69

of metacognition(Alexander and Judy, 1988;Clarke, 1992;Lajoie,1993). Characteristic

Data

Knowledge domain

Rules and principies

THE NEED FOR ILL-DEFINED PROBLEM TYPES

The tendency toward repetitive practice at routine problem solvingas the traditional approach in science teaching has been highlightedpreviously. The need for science graduales to be able to solve thecomplex, interdisciplinary, real-world problems they will face whenthey enter the workforce was also aI1uded to by ZoI1er (2000). Thetypes of problems students are asked to salve must be a major consid-eration in developing transferable problem-solving ski11s.

Io ibis end, Jonassen (2000) has provided a classification of prob-lem types and their characteristics. Ihe problem types identifiedinclude logical, algorithmic, rule-using, decision-making, trouble-shooting, diagnosis, case analysis, design, dilemmas, and stories. Forthe purposes here it is sufficient to consider a range of problems vary-ing alonga continuumfrom weI1-definedto ill-defined.We11-definedproblems are those that are typically seen at the end of chapters in stu-dent textbooks, and are designed for self-study and reinforcement ofkey concepts. Usua11ythey present all the elements of the situation;demand a limited number of skills, ruIes, and principIes; and require acorrect solution through a prescribed solution process in a well-de-fined dornajo of knowledge. In contrast, i11-definedproblems requirestudents to interpret some of the problem elements and may possessmultiple solutions or approaches. As it may be nuclear which rules orprincipIes are necessary for a solution, the learner needs to think stra-tegically, employ metacognitive skills, and defend bis or her solution.In general, any particular problem would fa11somewhere on a spec-trum ranging between the two extreme problem types. Table 4.1 indi-cates some differences in the characteristics of well- and i11-defined

problems. The deve1opment of high-1evel problem-solving skiI1sre-quires that students be given open-ended or ill-defined prob1ems.

In the specific context of ca1culation-type chemistry prob1ems,Johnstone (1998) has classified problems along three dimensions:whether data are complete or not, whether the method is familiar ornot, whether the solution or goals are given or oren. This classifica-tion results in eight types of problems, increasing in difficulty as databecome incomplete, the method becomes unfamiliar, and the solution

Solution process

Answer

Well-defined problem III-defined problem

Complete Incompleteornot givenWell-defined III-defined

Limited rules and principies Uncertainty about conceptsin organized arrangement and principies necessary for

solution

Unfamiliar; no explicitmeans for action

Uncertain, multiple or nosolution; need to make judg-ments and evaluations

Familiar; knowable, com-prehensible method

Clear goal, convergent;possess a correGí answer

becomes oren. The advantages to the students in solving each type ofproblem are also outlined, with higher-order cognitive skills beingdevelopedas more i11-definedproblems are tackled. .

Some specific examples offirst-year chemistry problems follow toi1lustrate a range of problem types.

Problem 1: Excited hydrogen atoms produce many spectrallines. One series of lines, caI1ed the Pfund series, occurs intheinfrared region. It results when an electron changes fromhigher levels to a level with n =5. Calculate the wavelength ofthe lowest energy Hne of ibis series.

Problem 2: Exposure to high clases of microwave radiation cancause damage. Estimate how many photons with A = 12 cm

must be absorbed to raise the temperature of your eye by3.0°C. How long would it take for your eye to be heated ibismuch if it was placed near a typical but detective microwaveoyen in such a position that it received one-hundredth of thepower output of the oyen?

Problem 3: We view visible light with OuTeyes every day with-out any ill effects. However, exposing OuTunprotected skin,which is much less "sensitive" iban OUTeyes, to a day on thebeach can have very painful consequences. What is happen-ing in each of these circumstances at the molecular level?Give as many details of the chemistry as possible and includequantitative calculations, if possib1e, to back up the explana-tions.

70 TEACHINGIN THESCIENCES

Problem 1 is a typical end-of-chapter exercise, with one correct an-swer, which can be obtained through application of an algorithm. Allof the data (except for values of fundamental constants) required forthe caIculation are supplied. The problem has liUle relation to anyphenomena a student can easily relate lo. To answer Problem 2, stu-dents will need to determine what extra data are required for a solu-tion and then search to find this information. (Data on the specificheat capacity of fue eJe and the power level of a microwave oyen areneeded and might have to be estimated by some sensible approxima-tion.) Reasonable values for the answers are required, rather than ex-act numbers. The setting of fue problem relates more to student expe-rience than Problem 1. Problem 3 also relates to students' everydayexperience. The way students answer the question depends muchmore on what they decide to do and research. It is an open-endedquestion and cIearly no one answer is correct. This question has beenused as a collaborative question, where students work as a team toproduce their answer.

It is not suggested that practicing routine, well-defined problemsdoes not benefit student learning. Recall that fue analysis of Taconis,Ferguson-Hessler, and Broekkamp (2001) showed that well-struc-tured dornajo knowledge is important in problem solving. Practice insolving welI-defined problems, in a properIy supportive learning en-vironment, can help this development. At first-year level, when stu-dents are still struggling to build a well-organized knowledge base intheir science subjects, it is important to allow students sufficient timeand practice with concepts in arder to develop meaningful under-standing. It is all toa easy to borden students with more and more con-tent, facts, and inert knowledge. Mayer (1997) suggests that it is amistake to believe in the idea of prior automatization, namely, thatstudents can only develop higher-order thinkingskills after they havemastered the prerequisite lower-order skilIs. Hence it is imperativethat students are given the opportunity to tackle real-world problemsand not simple routine exercises. There is evidence that further skilIsare required for success in ilI-defined problem solving. Shin, Jonas-seo, and McGee (2003) have found that dornajo knowledge and rea-soning skills are significant predictors of well-defined prob1em-solvingscores, whereas regulation of cognition and attitudes toward scienceare additional significant predictors of problem-solving success forill-defined problems.

Developing Metacognitive and Problem-Solving Skills

DESIGN OF TECHNOLOGY-SUPPORTEDMETACOGNITJVE TRAINING

71

¡I¡I

rl'

ii

How can problem-solving skills and instructional design princi-pIes develop metacognitive awareness be implemented in a techno-logical environment? Jonassen (1997) has proposed different in-structional design models for learning well-defined and ill-defined

problem solving. For weIl-defined problems, six steps are suggested:

1. Review prerequisite concepts, roles, and principIes.2. Present a conceptual model of the problem dornajo.

3. Model problem-solving performance with worked examples.4. Present practice problems.5. SuPpOrt the search for solutions.6. Reflect on the problem and the solution.

Different steps are suggested for ill-defined problems:

1. Articulatefueproblem context.2. Introduceproblem constraints.3. Locate, select, and developcases for learners.4. Supportknowledge-baseconstruction.5. Supportargumentconstruction.6. Assessproblem solutions.

Considering more specificaIly the development of higher-order

cognitive skills, metacognition, and reflection, three important impli-cations of social constructivist theory should be built in to fue learn-ing environment from the beginning. First, in arder to foster reflectivethinking, students need multiple sources offeedback on tbeir under-standing gained through social interactions. Second, reflective think-

ing will most likely occur in situations where problems are complexand meaningful to the student. Third, reflective thinking requires stu-dents to organize, monitor, and evaluate their thinking and learning tocome to a deeper understanding of their own processes of learning(Elen and Lowyck, 1999). In accordance with these principIes, Lin

72 TEACHING IN THE SCIENCES Developing Metacognitive and Problem-Solving Skills 73

and colleagues (1999) prescribe tour features that can provide scaf-folds to enhance reflection in the technology-based environment:

. Process displays: where students are explicitly shown what they

are doing in performing a task. Process prompts: where students are asked to explain what they

are doing at different stages throughout their problem-solvingprocedure. Process modeling: where students have access to databases andaudio or video displays explaining what, how, and why otherstudents and experts do what they do in solving a specificprob-lem

. Reflective social discourse: where students share their leamingexperiences and gajo feedback from a community of 1earners,for example, through an online discussion space

In a study conducted by Lin and Lehman (1999), biology studentsworked in a computer-based biology simulation leaming environ-ment, designing and conducting experiments involving the control ofvariables. The study investigated the use of explicit prompts toengage students in metacognitive thinking and problem solving in-volving control of variables. Students were given different types ofprompts: reasoned justification, rule-based, or emotion-focused. Qual-itative data showed that the reasoned justification prompts directedstudents' attention to understanding when, why, and how to employexperimental design principIes and strategies, and ibis in tum helpedstudents to transfer their understanding to a novel problem (Lin andLehman, 1999).

METAHEAD:AN ONL/NE TUTOR/ALTO SUPPORT METACOGNlTION

Instructional designers and teachers need to be able to adapt thesesuggestions and principIes to the particular Ieaming environment be-ing built, based on their own expertise and judgment in what is in es-sence the solution of an iU-defined prob1em.

There are several examples of technology-supported approaches tometacognitive skills and prob1em solving. LUCID (Learning and Un-derstanding through Computer-based Interactive Discovery) is a newmodel for computer-assisted leaming workshops to promote studentengagement in the leaming process (Wolfskill and Hanson, 2001).LUCID provides students with an orientation for the leaming processand allows them to freely navigate through activities. Key questions are

employed to guide the exploration of interactivemodels and the develop-meDíof understanding, with instant multilevel feedback (for questionswith a single correct answer) to promote confidence whi1e developingproblem-solving skills. More complex questions network reportingand peer-assessment, promoting critical review and the achievement oíconsensos in a group. Io assist in reflection and self-assessment, per-formance distributions are provided on the quantity and quality of thework and reports of other teams. This example illustrates how informa-tion and computer technologies (ICT) can enhance learning activitiesand the importance of using sound pedagogical principIes to drive theimplementation of ICT.

The changed profile oí students currently entering universities inAustralia is creating pressure to change teaching practices. Many stu-dents are matUTeage and work part -time while studying, creating ademand for more courses and programs in the distance-educationmode. Many students have been away írom study for a number oíyears and may not have the well-developed learning skills requiredfor tertiary study and may have a limited background in science.Moreover, distance students may íeel a sense oí isolation in not beingable to work cooperatively with other students or to compare thequality oí their own work in relation to that oí others enrolled in theircourses. This requires university teachers to support leaming in whatis a new afea of education for many first-year students. The metAHEADtutorial, an online tutorial provided for both Do-campos and off-cam-pus students taking first-year biology, biophysics, or chemistry at theUniversity oí New England (UNE), was developed in ibis context.

There has been debate about the relative benefits oí general studyskills programs versus explicit skills training within subject teaching.Research indicates that the teaching of problem solving is best leamedwithin the subject dornajo, rather iban as a separate, decontextualizedsubject (Mayer, 1997). Research suggests that students are less likelyto be interested in skills-development programs unless they can seedirect application to the work they are doing in their subjects at the

74 TEACHINGIN THESCIENCESDevelopingMet(U;ognitiveand Problem-SolvingSkills 75

problem solving andmadeawareoftheirownproblem-solv-ing strategies.

Phase 6: Successful students are presented with further prob-lems in the topic arca to check whether they have transferredfue strategies leamed during Phases 3 and 4. If they have not,training continues.

Phase 7: Students are given the opportunity to reflect on theirproblem solving.

Phase 8: This final step involves a refinement of the training tocreate design guidelines for a problem-solving environmentin different subject arcas (biology, biophysics, and chemis-try) in arder to foster metacognition.

A flow chart diagram for these phases as applied in the metAHEADtutorial is shown in Figure 4.1.

The overall structure .of the tutorial comprises tour modules, asshown in Figure 4.2. In the introductory module, Module 1, studentsencounter some basic ideas abOlir learning and thinking, cognition,and metacognition. They algo have the opportunity to take short ques-tionnaires relating to problem solving and metacognition to helpthem assess their current level. Module 2 introduces students to con-

time they are engaged in a tutorial. The aim then was to maximize theattractiveness of the tutorial by highlighting the cornmonality ofskills across biology, chemistry, and biophysics and starting with ac-tivities very similar to students' assignment tasks for various topics ofstudy.

The objectives, in designing the tutoríal, were as follows:

. To supportthe developmentof the metacognitiveskillsandhab-its of refleetion,which are essentialto effeetiveproblemsolvingin the sciences. To foster students' problem-solving skills in first-year seienee,utilizing the cornmunieative and supportive features of a tech-nology-based environment. To apply constructivist instructional design principIes that cancontribute to the development of an online environment to fostermetacognition

Based on the current research on metacognitive training, a scheme forthe development of metacognitive skills for scienee students that in-volved eight phases was adopted. The environment for metaeognitivedevelopment utilized Web-based learning activities to engage learn-ers in actual problem solving and reflection on their own problem-resolution strategies.

Phase5Intervention

Phase1: The eoneept of metacognitionis operationalized.Forthe problem in question, students need to become aware ofthe problem-solving processes involved. For example, thisrequires analysis of the question, planning a solution, and se-leetion of strategies and self-monitoring skills that can beused.

Phase2: Tbis phaseinvolvesthe design of the problemenviron-ment. For particular probIems in a topic in biophysics, for ex-ample, examine the different ways in which an expert and anoviee student rnight answer the problem.

Phase 3: The problem is presented to the student.Phase 4: Student responses are monitored to decide if any

intervention (Phase 5) is required.Phase 5: This step presents students with a seenario or problem

where they are assisted in the proeesses and proeedures of

~No

Phase1Operationalize

concept

Phase2Problem

environment

Phase3Student

doestaskNo

IVes

Phase8Refinetraining

Phase 7Student

reflectionVes

FIGURE 4.1. Phases tor Metacognitive Training

76

Module 2Coneepl maps

BiologyTopie1,

Topie 2, ele.

TEACHINGIN THESCIENCES

Ves

Modl¡le 3Deseriplions and

explanalions

Biophysics

Developing Metocognitive andProblem-Solving Skills 77

No

lem, being presented with a variety of prompts and questions abolirthe processes they engage in during their solution process. In thisway, students can be exposed to a number of process displays andprompts. They have an opportunity to víew answers given by otherstudents, rangíng from peor to very good answers. A model answerfrom the lecturer is also availab1e. All these answers are cornmented

upon and students may also listen to audio segments or video clips ofother students and lecturers as they worked on the problems. In thisway process modeling is provided.

For Selle questions, students are asked to collaborate with eachother to build a cornmunal answer on the bulletin board. There are op-portunities for reflective social discourse through discussion of par-ticular problems and also more general issues in bulletin board topics.Many of the same ideas are being applied, discussed, and retlectedupon in different questions on different topics and subjects, which arebelieved to assist in transfer of the skills developed to broader areas ofapplication. Throughout the tutoríal students are prompted to answerquestions and make notes in an online logbook in arder to keep arecord of their strategies and skill development.

Module1Inlroduelion lo

Ihinkingand learning

Module 4Problemsolving

Evaluation o/ metAHEAD

Chemistry

Oliver (1999) has remarked that while a range of methodologiesfor evaluation of ICT learning environments exists, each may be re-stricted in its use and in the range of situations to which it can be ap-plied. An evaluarían approach was sought that was broad and flexibleenough to suit the situation, namely, the introduction of an innovativeresource within a university context. The Oren University model, asdescribed by Jones, Tosunoglu, and Ross (1996), ís a useful frame-work. This approach recuses on three main themes: context, ínterac-tíon, and outcomes, as illustrated in Table 4.2. From the outset, datahave been gathered on the design and use of metAHEAD from pilarstudies, practitioners' opinions, instructional designers, and academicstaff.

The gatheríng of a range of data for the evaluarían of meta-cognition was facilitated by metAHEAD being Web-based. Data col-lected online included a self-rating quiz, taken at the beginning and atthe end of semester; students' self-predictions of success with partic-ular problems and retlection upon these on completion of the prob-

FIGURE 4.2. Modular Structure of metAHEAD Online Tutorial

cept mapping in each of the subject afeas. In the other modules stu-dents start with the typical exercises they come across in the study oftheir subjects, leading to metacognitive skills development as theytackle further problems. These modules have similar parallel pathsfor each subject; Module 3 considers tasks involving explanationsand descriptions, while Module 4 involves caIculations-type prob-lems. For a more detailed tlow chart of the path through a module, see

Hol1ingworth and McLoughlin (2001).When students lag on to the tutoríal, they are asked to choose a

subject and topic on which to work. They are asked to salve the prob-

78 TEACHINGIN THESCIENCES

TABLE4.2.Featuresof the EvaluationApproachAdopted

Feature

Rationale

Context

NeedformetAHEAD atUNE; currículumcontext

Designers aim;principies under-pinning design;pressing itera-tions

Interviews; writ-ten records

Interactions Outcomes

Need to look at stu- Learning outcomes; problem-dent interactions solving outcomes; changes ofwith the resource perception and attitude must

be considered

Measures of effective prob-lem solving, changes in atti-tu de, strategy, perception ofself

Data Records of stu-dents' interactions,student diaries,and online logs

Methods Observations, vid- Focus groups, tests, andeos, diaries, com- questionnairesputer records,product data gen-erated by students

km; online logbook enfríes,inc1udingstudents'noteson strategyuseand actualprob1emsolutions; and bul1etinboard discussions.Thesedata werecomplementedby furtherdatagatheredfrom small face-to-face focus group discussions.

Data have been gathered from focus groups giving feedback ontechnical issues and development of problem-solving skills. Twomajo conc1usionshave been gained from Ibis evaluarían.First, stu-dents have greatly appreciated the availability of other students' an-swers and particularly the comments on them. This has allowed stu-dents to pul their own answers joto a better perspective and to gajo aclearer idea of what the lecturer expects. (The fact that assessmentand lecturers' expectations drive the majority of students cannot beignored.) Access to alternative solutions provides students with pro-cess models and supports reflection by the students on many aspectsof prob1em solving. Coincidentally, other student answers help somestudents to feel more of a part of a group and that they are "not so stu-

pid after all," which can affect student motivation to succeed alongwith others. Students have noted that their motivation often parallelstheir success in the subject. Becoming part of a community of stu-dents working in a course is particularIy important for distance-education students, who may otherwise be rather isolated. The abilityto discuss issues on the bulletin board has been he1pful for these stu-dents.

,

I

I¡

Developing Metacognitive and Problem-Solving Skills 79

Second, students have mentioned that planning and analysis ofproblems, whereby parts can be tackIed step by step, has been helpfulto them. Becoming more aware of and practicing such skills has beenbeneficial for students who in the past may have been intimidated bya problem and may have given up too soon when it seemed too diffi-cult. Although some models and coaching online have been provided,students arrear to need a more staged approach, with step-by-stepexamples and support. This "cognitive apprenticeship" approach is intune with the constructivist principIes underpinning the design ofmetAHEAD. Box 4.1 shows a summary of student comments gainedfrom the evaluation.

Technical difficu16es present the greatest obstacles to the ease ofuse of the tutoríal at Ibis time. To be able to solve prob1ems, students

~need to be able to draw diagrams (in all subjects) and use mathemati- ~cal formulas and equations (particularly for chernistry and biophys- i tics). To carry out these tasks in any computer environment requires a Iconsiderable learning curve for most students at present. Moreover,while students can easily solve problems on paree employing dia- U.f\grams aml/orformulas,the online logbook,where they keep arecord 1'2of their progress over the semester, can handle plain text only. Theability to keep an adequate record of a student's progress over a se-mes ter in an online formal is a majar design feature if a more student-friendly, computerized environment is to be provided.

iIT.CONCLUSION l k

.roQ'I

LI L) ;J ..;,J

Students' metacognítiveskills can be developed significantlybytaking a proactive approach and by designing learning environmentsspecifically for problem solving and the development of metacog-nítion. The choice of problems students are exposed to is vitally im-portant forthe development ofhigher-Ievel cognitive skills. Leamingenvironments need to be developed in contexts that engage studentsin self-monitoríng their problem-solving approaches in scenarioswhere they can uItimately use that knowledge. Tbis requires creatingreal-life anchars for development of problem-solving skillS and en-abling students to explore, test, and review their own strategies.

These leaming environments do not necessarily need to be com-puter-based. At the mament the push for ICT-supported leaming

80 TEACH/NG IN THE SCIENCES Developing Metacognitive arul Problem-Solving Skills 81

BOX4.1.StudentCommentsGainedfromEvaluationof metAHEAD

situations is still far superior to the best computer tutors. It will re-quire that much more of the tacit knowledge of expert teachers bemade explicit and then incorporated joto the knowledge databases ofintelligent computer tutors before ICT can seriously replace face-to-face teaching. Significant technical difficulties algo limit the conve-nience of using computer environments for writing out and depictingsolutions to the sorts of problems students tackle in tertiary science.

The development of technology-supported resources is costly andtime consuming, and in-depth evaluation is essential not only for de-sigo but algo for development and implementation of such programs.An initial evaluation has shown both positive and negative aspects ofthe metAHEAD tutorial, and insights gained will be used to improvethe design of the resource and its capacity to support problem solv-ing. The choice of evaluation approach has been effective and worth-while, providing valuable data on context, interactions, and outcomesof the tutoría!. Continuing evaluation will assíst in further refinement,tailoring the tutorial to student needs and improving its relatedness tothe support oí metacognition in tertiary-Ievel science studies.

Comments on what students gained from metAHEAD

. Helps think about level of confidence and predict degree ofsuccess.. Helps motivation somewhat.Helps withways to understand the question better in arder totackle it successfully.Helps with breaking up tasks-something they didn't com-monly consider explicitly.Talkingto other students is helpful, hear other views.

. Getting feedback from lecturers, hearing them explain andcomment on lecturers' answers.. Appreciated other students' answers, so they could comparetheir own answers. Lecturers' "metacomments" on these wereimportant.. Other students' answers also show other ways of solvingproblems.. Information about learning and thinking processes-e.g.,chunking information.

REFERENCES

Commentson limitationsof metAHEAD

. Can't replaceface to face.Needpersonalinstructorformoti-valían.

. Needs to help students with more transfer to new problems.. Needs more models and demonstrations of how to workthrough problems step by step.. Logbook tool has great technical deficiencies, particularlyforscience answers involvingformulas and equations.. Audioand video quality needs to be better.

Alexander, PA and Iudy, IJ. (1988). The interaction of domain-specific and strate-gic knowledge in academic performance. Review of Educational Research,58(4): 375-404.

Boekaerts, M., Pintrich, P.R, and Zeidner, M. (2000). Handbook of self-regulation.San Diego, CA: Academic Press.

Clarke, RE. (1992). Facilitating domain general problemsolving: Computer cogni-tive processes and instruction. In E. De Corte, M.e. Linn, and H. Mandl (Eds.),Computer-based learning environments and problem solving (pp. 265-283).Berlin: Springer Verlag.

Elen, 1. and Lowyck, 1. (1999). Metacognitive instructional knowledge: Cognitivemediarían and instructional designoJournal of Structural Leaming and lnteUi-gentSystems, 13(3): 145-169.

Everson, H.T. and Tobias, S. (1998). The ability to estímate knowledge and perfor-mance in college: A metacognitive analysis.lnstructional Science, 26(1-2): 65-79.

Gabel, D. (Ed.) (1994).Handbook ofresearch on science teaching and leaming.New York: Macmillan Publishing Company.

Gourgey, A.E (1998). Metacognition in basic skills instruction. lnstructional Sci-elIGe,26(1-2): 81-96.

Gredler,M.E. (1997).Learning and instruction: Theory into practice. Merrill, NI:Prentice-Hall.

comes about partly through a need to accornmodate distance-educa-tion students, as well as the need to teach larger student numbers with

decreasing staff numbers. In fact, using ICT to assist the developmentof higher-order cognitive skills represents a very significant chal-lenge. A face-to-face encounter of a learner with an experienced per-sonal bullan instructor who has expert subject knowledge and peda-

gogical knowledge bases and is able to react flexibly to different

82 TEACH/NG IN THE SCIENCESDeveloping Metocognitive and Problem-Solving Skills 83

Schraw, G. (1998). Promoting general metacognitive awareness. lnsrructional Sci-ence, 26(1-2): 113-125.

Shin, N., Jonassen, D.H., and McGee, S. (2003). Predictors of well-structured andill-structured problem solving in an astronomy simulation. Joumal of Researchin Science Teaching, 40(1): 6-33.

Taconis, R, Ferguson-Hessler, M.G.M., and Broekkamp, H. (2001). Teaching sci-ence problem solving: An overview of experimental work. Joumal of Researchin Science Teaching, 38(4): 442-468.

Volet, S.E. (1991). Modeling and coaching of relevant metacognitive strategies forenhancing university students' learning. Learning and lnstruction, 1: 319-336.

Volet, S.E., McGill, T, and Pears, H. (1995). Implementing process-based instruc-

tion in regular university teaching: Conceptual, methodological and practical is-sues. European Journal of Psychology of Educarion, 10: 385-400.

Wolfskill, T. and Hanson, D. (2001). LUCID: A new model of computer-assistedlearning. Joumal ofChemical Educarion, 78(10): 1417-1424.

Zeegers, P., Martin, L, and Martin, C. (1998). Using learning to learn strategies toenhance student self-regulated leaming in first-year chemistry. Papel presentedat the Third Pacific Rim Conference, Auckland, New Zealand, July 5-8.

Zoller, U. (2000). Teaching tomorrow's college science courses-Are we getting itright? Joumal ofCollege Science Teaching (May): 409-414.

Hatano, G. and Inagaki, K. (1986). Two courses of expertise. In H.AH. Stevensonand K. Hakuta (Eds.), Child developmenr and educarían in lapan (pp. 262-272).New York: Freeman.

Hobden, P. (1998). The Tole of routine problem tasks in science teaching. In BJ.Fraser and K.G. Tobin (Eds.), lntemational handbook of science educarían, Vol-lime 1, Part 1 (pp. 219-231). Dordrecht, the Netherlands: Kluwer AcademicPublishers.

Hollingworth, RW. and McLoughlin, C. (2001). Developing science students'metacognitive problem-solving skills online. Australian loumal of Educational

Technology, 17: 50-63.Johnstone, AH. (1998). Leaming through problem solving. In D. Rafferty and S.

Sleigh (Eds.), Problem solving in analytical chemistry (pp. v-viii). London: The. Royal Society of Chemistry.Jonassen, D. (1997). Instructional design models for well-structured and ill-struc-

tured problem-solving leaming outcomes. Educational Technology Researchand Development, 45(1): 65-94.

Jonassen, D. (2000).Towarda design theoryof problem solving.Educational Tech-nology Researchand Development,48(4): 63-85.

Jonassen, D. (2002). Learning to solve problems online. In G. Glass and C. Vrasidas(Eds.), Currentperspectives on applied informarían technologies. Volume 1:Distance leaming (pp. 75-98).Macomb:Center for the Application ofInforma-tion Technologies, University of Western Illinois.

Jones, AE.S., Tosunoglu, C., and Ross, S. (1996). Evaluatíng CAL at the Oren Uní-versity: 15 years oo. Computersin Education, 26(1):5-15.

Lajoie, S.P. (1993). Computer environments as cognitive tools for enhancing learn-ing. In S.P. Lajoie and SJ. Derry (Eds.), Computers as cognitive tools (pp. 261-288). Hillsdale, NJ: Lawrence Erlbaum.

Lin, X., Hmelo, C., Kinzer, C.K., and Secules, TJ. (1999). Designing technology tosupport reflection. Educational TechnologyResearchand Development, 47(3):43-62.

Lin, X. and Lehman, J.D. (1999). Supporting learning of variable control in a com-puter-based biology environment: Effects of promptíng college students to re-flect on their own thinking.loumal of Research in ScienceTeaching,36(7): 837-858.

Lowe, I. (1999). Education for a rapidly changing future. Available online at<http://www.aisq.qld.edu.au/publications>.

Masui, C. and De Corte, E. (1999). Enhancing learning and problem-solving skills:Orienting and self-judging, two powerful and trainable learning tools. Learningandlnstruction, 9(6): 517-542.

Mayer, RE. (1997). Incorporatingproblemsolving into secondary schoolcurricula.In G.D. Phye (Ed.), Handbook of academic leaming: Construction ofknowledge(pp. 474-492). San Diego, CA: Academic.

Metcalfe,1. and Shimamura, A.P. (1994). Metacognition: Knowing abolIr knowing.Cambridge, MA: MIT Press.

Oliver, M. (1999). ELT report 1. A framework for evaluating the use of educationaltechnology. Available online at <http://www.unl.ac.uk>.