J. EDUCATIONAL COMPUTING RESEARCH, Vol. 31(3) 215-245, 2004

CAN STUDENTS COLLABORATIVELY USE HYPERMEDIA

TO LEARN SCIENCE? THE DYNAMICS OF SELF-

AND OTHER-REGULATORY PROCESSES

IN AN ECOLOGY CLASSROOM*

ROGER AZEVEDO

FIELDING I. WINTERS

DANIEL C. MOOS

University of Maryland

ABSTRACT

This classroom study examined the role of low-achieving students’ self-

regulated learning (SRL) behaviors and their teacher’s scaffolding of SRL

while using a Web-based water quality simulation environment to learn about

ecological systems. Forty-nine 11th and 12th grade students learned about

ecology and the effects of land use on water quality by collaboratively

using the RiverWebSM water quality simulation (WQS) during a two-week

curriculum on environmental science. The students’ emerging understanding

was assessed using pretest and posttest scores. Students’ self-regulatory

behaviors and teacher’s scaffolding of SRL were assessed through an analysis

of their discourse during several collaborative problem-solving episodes.

Findings indicate that students learned significantly more about ecology

after working collaboratively with the WQS. However, these learning

gains were quite small and were related to the self-regulatory behaviors

observed in the dyads and their teacher’s scaffolding and instruction.

Analyses of video data indicate that a large amount of time was spent by the

dyads and teacher in using only a few strategies, while very little time was

spent on planning, monitoring, and handling task difficulties and demands.

Further analysis revealed that both the dyads and teacher were using low-level

*This research was supported by funding from the National Science Foundation (REC#0133346)

awarded to the first author.

215

� 2004, Baywood Publishing Co., Inc.

strategies (e.g., following procedural tasks, evaluating the content, searching,

and selecting informational sources in the WQS) to learn about the topic. Our

results provide a valuable initial characterization of the complexity of self-

and other-regulated learning in a complex, dynamic, technology-enhanced,

student-centered science classroom. We discuss how the results will be used

to inform the design of computers as MetaCognitive tools designed to foster

students’ learning of conceptually challenging science topics.

INTRODUCTION

Educational researchers have successfully used student-centered methods to

enhance students’ understanding of science with computer-based learning

environments (CBLEs) (e.g., Azevedo, Verona, & Cromley, 2001; Biswas,

Schwartz, Bransford, & CTGV, 2001; Clark & Linn, 2003; Hickey, Kindfield,

Horowitz, & Christie, 2003; Kozma, Chin, Russell, & Marx, 2000; Lajoie &

Azevedo, 2000; Lajoie, Guerrera, Munsie, & Lavigne, 2001; Reiser et al., 2001;

Slotta & Linn, 2000; Suthers & Hundhausen, 2003; Vye et al., 1998; White,

Shimoda, & Frederiksen, 2000). These studies, however, also provide evidence

that students have difficulties regulating certain aspects of their learning when

they use CBLEs to learn about complex and challenging science topics such as the

circulatory system, Newtonian physics, and genetics (e.g., Hickey et al., 2003;

Jacobson & Archodidou, 2000). More specifically, research shows that students

have difficulties regulating several aspects of their learning, including their

cognitive resources (e.g., activating prior knowledge), motivation and affect

(e.g., task value, self-efficacy, interest in the topic), behavior (e.g., engaging in

help-seeking behavior), and the instructional context (e.g., negotiating task struc-

ture and dealing with dynamically changing structure of the instructional context).

While the majority of these studies have examined specific aspects of students’

self-regulated learning (SRL) such as planning, metacognitive monitoring, and

reflection in isolation (e.g., Lajoie et al., 2001), most have not examined the

complex nature of the dynamics between the phases of SRL and how this com-

plexity may affect how students regulate their learning of science topics in

student-centered classrooms using CBLEs.

It is therefore critical that we understand how students working collaboratively

regulate their own learning as well as how teachers facilitate students’ SRL

by using different scaffolding and instructional approaches in student-centered,

technology-rich science classrooms. To that end, we have adopted and extended

existing models of SRL (see Boekaerts, Pintrich, & Zeidner, 2000; Zimmerman

& Schunk, 2001 for extensive reviews) and have used them as a lens to examine

the complex interactions between students’ self-regulatory behavior and teachers’

scaffolding of their students’ SRL when using CBLEs collaboratively in the

science classroom for extended periods of time.

216 / AZEVEDO, WINTERS AND MOOS

Theoretical Framework: Self-Regulated Learning

By adopting SRL as framework, we make certain theoretical and empirically-

based assumptions about learning. Self-regulated learners are active learners who

efficiently manage their own learning in many different ways (Boekaerts et al.,

2000; Butler, 1998; Paris & Paris, 2001; Perry, 2002; Winne, 2001; Winne &

Hadwin, 1998; Winne & Perry, 2000; Zimmerman & Schunk, 2001). Students

are self-regulating to the degree that they are cognitively, motivationally, and

behaviorally active participants in their learning process (Zimmerman, 2001).

These students generate the thoughts, feelings, and actions necessary to attain their

learning goals. Models of SRL describe a recursive cycle of cognitive activities

central to learning and knowledge construction activities (e.g., Pintrich, 2000;

Schunk, 2001; Winne, 2001; Winne & Hadwin, 1998; Zimmerman, 2001).

Several models of SRL posit that SRL is an active, constructive process

whereby learners set goals for their learning and then attempt to monitor, regulate,

and control their cognition, motivation, and behavior in the service of those goals

(see Pintrich, 2000; Zimmerman, 2001). SRL is guided and constrained by both

personal characteristics and the contextual features of the environment (Pintrich,

2000). Thus, these models offer a comprehensive framework with which to

examine how students learn and how they adapt during learning with hypermedia.

Recent research has begun to examine the role of students’ ability to regulate

several aspects of their cognition, motivation, and behavior during learning of

complex science topics with hypermedia in laboratory experiments and in tutoring

situations (e.g., Azevedo, Cromley, & Seibert, 2004; Azevedo, Guthrie, & Seibert,

2004). Thus far, this research has demonstrated that students have difficulties

benefiting from hypermedia environments because they fail to engage in key

mechanisms related to regulating their learning with hypermedia.

Self-Regulated Learning with CBLEs

In CBLEs such as Web-based hypermedia learning environments, students

are given access to a wide range of information represented as text, graphics,

animation, audio, and video, which is structured in a non-linear fashion (Jonassen,

1996). Learning in such an environment requires a learner to regulate his or

her learning; that is, to make decisions about what to learn, how to learn it, how

much to learn, how much time to spend on it, how to access other instructional

materials, determine whether he or she understands the material, when to abandon

and modify plans and strategies, and to increase effort (Shapiro & Niederhauser,

2004; Williams, 1996; Winne & Hadwin, 1998). Specifically, students need to

analyze the learning situation, set meaningful learning goals, determine which

strategies to use, assess whether the strategies are effective in meeting the learn-

ing goal, evaluate their emerging understanding of the topic, and determine

whether the learning strategy is effective for a given learning goal. They need

to monitor their understanding and modify their plans, goals, strategies, and

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 217

effort in relation to contextual conditions (e.g., cognitive, motivational, and task

conditions) (Azevedo, Cromley, & Seibert, 2004). Further, depending on the

learning task, they may need to reflect on the learning episode and modify their

existing understanding of the topic (Winne, 2001; Winne & Hadwin, 1998).

Therefore, despite the potential for fostering learning, hypermedia environments

are proving to be ineffective unless learners can regulate their learning. Being

able to regulate one’s own learning is therefore extremely important in enhancing

one’s ability to learn about complex topics with hypermedia. Researchers have

recently begun to identify several problems associated with learners’ inability

to regulate their learning with hypermedia (Azevedo, Cromley, Winters, Xu,

& Iny, 2003). The next step in this line of research is to extend these results

by focusing on whether students can collaboratively use Web-based hypermedia

learning environment to learn about the effects of land use on water quality

indicators in a student-centered classroom by focusing on the complex interactions

between students’ SRL.

Self-Regulated Learning with CBLEs in the

Classroom: Complex Contextual Interactions

The importance of the classroom context for students’ self-regulated has been

well-established (e.g., Butler, 1998; Meyer & Turner, 2002; Paris & Paris, 2001;

Perry, 2002; Perry & VandeKamp, Mercer, & Nordby, 2002). Current research

has focused on the importance of the classroom as a complex context within

which to study students’ SRL and the role of teachers (as external self-regulating

agents) who foster students’ understanding by using scaffolding or any other

instructional methods to support their learning (e.g., Rasku-Puttonen, Etelapelto,

Arvaja, & Hakkinen, 2003). However, what has not been investigated is the

complex dynamics between students’ SRL while working collaboratively and

using a CBLE to learn about a challenging science topic, and their teacher’s

scaffolding of their SRL in the classroom.

Meyer and Turner (2002) emphasized that in a classroom context there is a

shared responsibility among teachers and students for establishing and main-

taining relationships as they coordinate multiple goals through available oppor-

tunities and scaffolding (McCaslin & Hickey, 2001). Scaffolding is an instruc-

tional process in which the teacher supports students cognitively, motivationally,

and emotionally in learning while helping further develop their autonomy

(Meyer & Turner, 2002). During instructional scaffolding, the teacher supports the

students’ self-regulation, as needed, in three ways: a) helping students build

competence through increased understanding; b) engaging students in learning

while supporting their socioemotional needs; and c) helping students build and

exercise autonomy as learners. The interactions between a teacher and a class full

of students with varying needs and abilities to regulate their learning while

working collaboratively with a CBLE to learn about a challenging science topic

218 / AZEVEDO, WINTERS AND MOOS

pose theoretical and methodological challenges which need to be addressed in

order to study whether students working collaboratively can use CBLE to learn

about challenging topics.

In this study, we address the question of whether academically low-achieving

students, who have failed several science classes in high school, can regulate their

learning of conceptually challenging science topics such as the effects of land use

on water quality in a student-centered technology-enhanced science classroom.

Our study contributes to a large body of literature on fostering students’ learning of

conceptually challenging science topics with CBLEs in student-centered class-

rooms. It examines low-achieving students’ SRL and the teacher’s scaffolding

of students’ SRL in a classroom setting. It converges product data (i.e., learning

gains) and process data (i.e., time and frequency of use of several SRL variables

used by students working collaboratively, as well as teacher scaffolding and

instructional moves). This classroom study was conducted to investigate the

following research questions: 1) Can low-achieving high school students effec-

tively use a water-quality simulation1 (WQS) collaboratively to learn about

water quality indicators and land use? 2) Do the students and teacher spend the

same amount of time using various regulatory behaviors to facilitate students’

conceptual understanding? and 3) Do the students and teacher differ in the

frequency of use of these self-regulatory behaviors to facilitate students’ con-

ceptual understanding?

METHOD

Participants

Forty-nine 11th and 12th grade high school students (22 girls and 27 boys) from

four ecology science classes volunteered to participate in this study. Ages ranged

from 16 to 18 years. The students came from diverse socioeconomic and racial

backgrounds. Forty-nine percent (n = 24) were African American, 41% (n = 20)

were Hispanic, 6% (n = 3) were Caucasian, and 4% (n = 2) were Asian. All of

these students were placed in an environmental science class after failing several

required science classes in high school. The students had limited exposure to

environmental science, and therefore low prior knowledge about it. In addition, we

observed that students lacked both general learning skills (e.g., comprehension

monitoring) and specific science-related skills (e.g., how to round numbers).

Data were collected in April and May 2003, one month prior to the end of

the school year.

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 219

1 The original designers and developers of the RiverWeb WQS labeled it as a simulation-based

environment, however, we consider it to be a Web-based hypermedia learning environment since it

provides students access to multiple representations of scientific information in multiple formats

including text, and diagrams, graphics in a non-linear fashion.

Research Design

We used a one-group pretest-posttest design (49 students). Video and audio

data were collected while the students worked in groups of two using the

RiverWeb Water Quality Simulator (WQS) to learn about ecological systems.

We collected process data on dyads of students from two of four classes taught

by two environmental science teachers.

Instructional Context

In a typical high school environmental science course, a unit is devoted to

water quality and the impact of land use on the pollutants that find their way

into streams, rivers, and eventually, the ocean. In order for students to understand

the differential impact of agricultural, commercial, urban, suburban, manufac-

turing, and foresting/lumbering land uses on water quality, students need to have

some kind of experience with each. An ideal project for students in such an

environmental science course would be to investigate how water quality changes

across different locales; however, that would require field trips that are both

expensive and time intensive and therefore often infeasible. The WQS is designed

to provide students with a simulated field trip through a prototypical watershed.

The RiverWebSM Water Quality

Simulation (WQS) Environment

The WQS is a Web-based environment (http://mvhs1.mbhs.edu/riverweb/

index.html) developed at Maryland Virtual High School and linked to the

RiverWebSM Program originating from the National Center for Supercomputing

Applications (NCSA). Targeted at the grades 8 through 12 science and math

curriculum, the WQS depicts the effects of various land uses on water quality in

an archetypal watershed. Students access WQS monitoring stations to explore how

different land uses (i.e., a pristine forest, agricultural, lumbering forest, residential

area, commercial/industrial area, wetlands, and urban area) affect water quality.

Each water monitoring station allows students to test for physical and chemical

characteristics of the tributary (termed indicators), such as total flow and nitrogen

concentration. By limiting each sub-watershed to one land use, the effect of that

land use can be seen on the quality of the water that students “test” within its

boundaries. The cumulative effect of the combined land use determines the water

quality shown by the indicator values found at a common river outflow.

After the user logs in, a map of the archetypal watershed appears. Water quality

monitoring stations located throughout the watershed are depicted on the map.

The user may click on the map to investigate any sub-watershed using the WQS

graph window. After clicking on a station, the student is presented with a default

graph that displays the variation of nitrogen over time in the top window and

precipitation over time in the bottom window. However, other indicators may

220 / AZEVEDO, WINTERS AND MOOS

be selected; for example, the student could compare different indicators at

the same station such as water temperature and toxins (as shown in Figure 1),

or could compare nitrogen concentration between two stations (e.g., residential

area vs. wetlands). In addition, by reducing the range of days for each graph,

the student can zoom in on a particular time period and investigate daily or

seasonal variations.

A tour, which may be selected at login, uses frames to combine the WQS with

instructions leading the user through most of the simulator’s capabilities. Students

can examine time series reports and scatterplots: using the time series graphs,

students are able to see the impact of seasonal changes on certain indicators;

using the scatterplots, students are able to discover correlations between variables,

such as the relationship between water temperature and dissolved oxygen. The

goal is to foster student understanding of the complex relationships among the

many land uses, indicators, and water quality in the watershed. To gain this

understanding, students are first given instruction in graph interpretation. They

then answer open-ended questions, using the WQS, which demand explanations

of concepts and relationships, and finally they construct concept maps.

Data Sources and Measures

Several data sources were used to obtain an in-depth understanding of students’

emerging understanding of science phenomena while using the WQS. A total of

6.12 hours of video and audio data were collected on four dyads in each of the two

science classes during all online science inquiry sessions with WQS. This allowed

for an in-depth analysis of students’ behavior as well as teacher instruction and

scaffolding while students engaged in collaborative science inquiry activities

with the WQS. The students’ emerging understanding and ability to regulate their

learning was assessed through an analysis of student and student-teacher inter-

actions during these activities. In addition to the video and audio data from student

pairs, we also collected students’ pretests and posttests.

The paper-and-pencil materials consisted of a pretest and a posttest, which were

constructed by a consulting science teacher based on information presented in

the WQS. The pretest and posttest, which were identical, included seven complex

questions which required students to: 1) demonstrate understanding of runoff

in relation to different water quality indicators and sources and effects of different

indicators; 2) describe and hypothesize daily and seasonal variations depicted

on several line diagrams; 3) determine relationships between different indicators

at different land use areas using scatter plots; and 4) construct a concept map

illustrating the relationships between water quality indicators and land use. The

range of total possible points for each test was 0–62.

The questions were designed to give each student an opportunity to demonstrate

his or her understanding of various issues related to water quality. Question 1

was designed to determine if students understood whether pollutants found at the

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 221

222 / AZEVEDO, WINTERS AND MOOS

Figure 1. RiverWeb interface illustrating a student comparing the

variation in water temperature (top window) and toxins (bottom window)

over time in the wetlands area (Station 5).

mouth of a river come from runoff from land uses areas upstream (range 0–5).

Question 2 was designed to assess whether students understood the importance of

trees as a stream buffer, particularly in developed areas (range 0–6). Question 3

was designed to assess whether students understood how land use can cause an

increase in runoff and, subsequently, more sediments and toxins in the water

(range 0–6). Question 4 was designed to assess whether students understood that

there is seasonal variation in these variables and why that might be (range 0–6).

Question 5 was designed to assess two concepts; each scored separately, whether

students understood that the scale on the y-axis of any graph must be included in

the interpretation of the graph, and that land usage affects nitrogen concentra-

tion (range 0–9). Question 6 also had two parts, scored separately, and was

designed to assess whether students understood that the scale on the x-axis of

any graph must also be included in the interpretation of the graph and that

both precipitation and land use affect runoff (range 0–8). In question 7, students

had to draw a concept map to show that they understood how various factors

affect water quality (range 0–22). With regard to format, questions 1, 2, and 3

were constructed-response questions. Questions 4 and 5 combined time series

graph interpretation and constructed response. Question 6 was a scatter plot graph

interpretation question, and question 7 asked the students to draw and label

a concept map.

Procedure

For this study, students in the environmental science course spent two weeks

using the WQS. For most of the students, this was their first experience learning

about water quality. On the first day of week one, the students took a pretest

to assess their understanding of water-quality indicators and land use. They

were then introduced to the concept of water quality and engaged in mini-labs

such as testing water for nitrogen or identifying the directionality and strength

of relationships between graphed water-quality indicators. During the end of

week one, the students were given a tour of the WQS and taught how to use the

various tools to analyze data. During week two, the students participated in three

“jigsaw” activities using the WQS. In Jigsaw 1, students used a time-series graph

to compare and contrast two variables, and then compared the indicators in

their land use area to those in the pristine forest. In Jigsaw 2, students used the

WQS to learn about two assigned indicators in each land use area. In Jigsaw 3,

students used scatter plots in the WQS to analyze the relationships of all the

different indicators for their specific land use area. The groups then reunited and

composed concept maps showing the links between the different indicators for

each area and presented their maps to the whole class. During the final class period

of week two, the teachers administered the posttest to all students.

Performance data, in the form of pretests and posttests, were collected for all

49 students. Process data, in the form of videotapes and audiotapes, were collected

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 223

on two separate occasions over the two-week period from a sample of eight

dyads from two classes. One consulting teacher acted as a complete participant

during the data collection period. She is an experienced environmental science

teacher who introduced the science activities to all of the students and provided

scaffolding to individual student pairs during activities. The regular classroom

teacher and the second author, a former science teacher, also provided scaf-

folding to students during their activities. The consulting teacher and the

two classroom teachers were not instructed to foster students’ SRL nor were

they asked to use specific SRL process (e.g., planning, creating sub-goals,

metacognitive monitoring) during this naturalistic classroom study. The first

author and several other graduate assistants acted as complete observers rather

than participants in the classrooms, remaining on the sidelines to take notes

and manage taping equipment, interacting minimally with the students and

teachers during class periods. Taping was done for whole class periods during

which the teachers moved in and out of interaction with individual groups

as they tackled the tasks in the jigsaw assignments. Data were therefore

gathered on groups of students while they worked both with and without teacher

assistance.

Coding and Scoring

In this section, we describe the scoring of the students’ answers to the pretests

and posttests, the coding scheme for student and teacher SRL behaviors, and

inter-rater agreement measures.

Students’ Answers to Pretest and

Posttest Questions

The second author and the consulting teacher constructed a rubric for scoring

the students’ responses to the pretest and posttest questions by initially scoring a

subset (30%) of all the pretests and posttests. The rubric was refined and the two

teachers scored all 98 tests individually using the revised rubric. Questions 1, 2, 3,

and 7 were each given a single score; question 5 had two parts so it was given

two scores; and questions 4 and 6 each had two parts so the two questions were

given four scores each. This yielded 14 separate scores for each student’s pretest

and posttest, which were combined to calculate each student’s overall (questions

1–7 on the pretest and posttest) score out of 62 for each test.

Students’ and Teacher’s Regulatory

Verbalizations and Behavior

The 6.12 hours (367 minutes) of audio and video tape recordings from 14

episodes provided the raw data for the analysis of students’ self-regulatory

224 / AZEVEDO, WINTERS AND MOOS

behavior and teacher’s scaffolding of SRL. Azevedo, Cromley, and Seibert’s

(2004) original model of SRL was modified and used for analyzing the students’

and teacher’s regulatory behaviors and adapted to fit the use of the WQS in the

classroom. Their model is based on several recent models of SRL (Pintrich, 2000;

Winne, 2001; Winne & Hadwin, 1998; Zimmerman, 2001). It includes key

elements of these models (i.e., Winne’s (2001) and Pintrich’s (2000) formulation

of self-regulation as a four-phase process) and extends these key elements to

capture the major phases of self-regulation in a complex technology-rich student-

centered science classroom.

The classes, descriptions and examples of the planning, monitoring, strategy

use, and task difficulty and demand variables used for coding the learners’ and

teacher’s regulatory behavior are described in the Appendix. The classes are:

a) planning and goal setting, activation of perceptions and knowledge of the

task and context, and the self in relationship to the task; b) monitoring processes

that represent metacognitive awareness of different aspects of the self, task, and

context; and c) efforts to control and regulate different aspects of the self, task,

and context; and d) various kinds of reactions and reflections on the self and

the task and/or context. The model also includes behavior of the teacher when

she provided scaffolding of and instruction in SRL-related behaviors in order

to enhance students’ understanding. Teacher scaffolding refers to any teacher-

initiated move that has the intention of facilitating students’ understanding

through various self-regulatory behaviors. Teacher instruction refers to any

teacher-initiated move that has the intention of specifically directing students’

SRL. For this study and in this context, the model was adapted to include verbal

as well as non-verbal behaviors specific to the tasks students engaged in while

learning with the WQS. As such, video data as well as audio data were used to

provide evidence of teacher and student SRL behaviors.

Inter-Rater Agreement

Each teacher scored each student’s questions on the pretest and posttest

(i.e., 7 questions with a total of 14 scores × 49 students × 2 tests = 1,372 individual

scores). Each teacher was instructed to independently code all 1,372 scores on

the pretests and posttests using a scoring rubric developed by both science

teachers. Then inter-rater agreement was established for the scoring of the

learners’ pretest and posttest answers by comparing the individual coding of

the high school teacher and the second author. There was agreement on 66 out

of 66 randomly selected pretests and posttests, yielding an inter-rater agreement

of 100%. For SRL behavior variables and time, inter-rater agreement was

established on 289 out of 305 coded segments (27% of codes), yielding inter-rater

agreement of 95%. Inconsistencies were resolved through discussion between

the experimenters.

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 225

RESULTS

Question 1: Can low-achieving high school students effectively

use a water-quality simulation (WQS) collaboratively to learn

about water quality indicators and land use?

We examined the shift in students’ learning (from pretest to posttest) by using

a paired samples t-test. There was a significant difference between the students’

mean pretest score and their mean posttest score (t[48] = –7.03, p < .001).

However, while there was a significant shift in the test scores, the students’ mean

score was low on both the pretest (M = 12.1%, SD = 4.5) and the posttest

(M = 19.1%, SD = 6.0).

Question 2: Do the students and teacher spend the same amount

of time using various regulatory behaviors to facilitate students’

conceptual understanding?

We examined whether the relatively small (yet significant) learning gains

from pretest to posttest were due to differences in the amount of time that both

students and teachers spent regulating students’ learning with the WQS across

each of the four SRL variables (planning, monitoring, strategy use, and handling

task difficulties and demands). A MANOVA was conducted to determine whether

students, teacher scaffolding, and teacher instruction differed in the amount of

time spent on each of these SRL variables (see Figure 2). The data met the

statistical assumptions for a MANOVA. There was a significant difference in

the mean time that learners spent regulating their own learning and the amount of

time the teacher spent on externally regulating the students by using scaffolding

and direct instruction (F[3, 117] = 19.32, p < .05). Overall, students spent sig-

nificantly more time regulating their own learning than the teacher spent on

regulating students’ learning either by instructing or scaffolding. There were

also significant differences in the amount of time students and teacher spent

using different strategies to regulate students’ learning. There were no differences

in the amount of time spent between planning, monitoring, and handling task

difficulties and demands. Multiple pairwise comparison tests found significant

differences between the time spent by students and the teacher on scaffolding

and instruction for four SRL variables. Learners spent significantly more time

using strategies to regulate their learning (F[2, 39] = 10.96, p < .05) than the

teacher did in providing either direct instruction or scaffolding students’ use

of strategies (Mean = 294.4 sec, SD = 236.3; Mean = 143.0 sec, SD = 78.0;

Mean = 39.0 sec, SD = 35.9, respectively). The teacher spent significantly

more time instructing students how to plan their learning (F[2, 39] = 10.26,

p < .05) than she did scaffolding their planning or that students spent planning

their own learning (Mean = 36.0 sec, SD = 5.5; Mean = 12.7 sec, SD = 5.5;

Mean = 1.8 sec, SD = 5.5, respectively). However, learners and teacher tended

226 / AZEVEDO, WINTERS AND MOOS

to spend the same amount of time regulating their own and her students’ moni-

toring activities (F[2, 39] = 0.20, p > .05), and handling task difficulties and

demands (F[2, 39] = 3.14, p > .05).

Question 3: Do the students and teacher differ in the

frequency of use of self-regulatory behaviors to facilitate

students’ conceptual understanding?

Third, we examined how learners and teachers regulated the students’ learning

of the ecology system using the WQS by calculating how often teacher instruction,

teacher scaffolding, and/or student behavior reflected a variable relating to one of

the four main SRL categories. We examined whether the frequency of use of each

of the variables relating to the four main SRL categories of planning, monitoring,

strategy use, and task difficulty were used significantly differently in teacher

instruction, teacher scaffolding, and student behavior. A total of 1,143 segments

were coded based on 6.12 hours (367 minutes) of video from the four dyads in each

of the two classes. Chi-square analyses revealed a significant difference between

the frequency of these variables used between students, teacher scaffolding, and

teacher instruction (�2[6, 1143] = 145.16, p < .001) (see Table 1).

Due to the overall significance, we probed further to examine what specific

SRL behaviors the students engaged in and teachers instructed or scaffolded

while learning with the WQS. In this section, we present a detailed account of

each of the specific SRL variables used by students and teachers to regulate the

students’ learning of ecology with the WQS.

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 227

Figure 2. Mean time spent by students and teacher on SRL behaviors.

Planning

Chi-square analyses demonstrated that students and teachers engaged in very

few behaviors associated with category of planning, accounting for only 4% of all

the codes. Under planning, students and teachers were coded for behaviors related

to planning (P), activating prior knowledge (PKA), stating goals (G), or recycling

a goal in working memory (RG). Students displayed no behavior associated with

prior knowledge activation or stating goals. Indeed, student moves were only

associated with 1%2 of codes in this category, due to a few instances of planning

228 / AZEVEDO, WINTERS AND MOOS

Table 1. Number and Proportion of Students’ Self-Regulated Learning

and Teachers’ Scaffolding of Students SRL Movesa

Student-

initiated

moves

Teacher-

instruction

moves

Teacher-

scaffolding

moves

SRL variable f % f % f %

Planning

(Planning, goals, prior knowledge

activation, and recycling goal in

working memory)

Monitoring

(Judgment of learning, partner

questioning, monitoring progress

toward goals, teacher questioning,

and feeling of knowing)

Strategy Use

(Follow directions, procedural,

content evaluation, search, and

selecting new informational

source)

Task Difficulty and Demands

(Time and effort planning and

help-seeking behavior)

Total

7

81

720

56

866

1%

9%

83%

7%

100%

33

21

131

20

205

16%

10%

64%

10%

100%

5

23

41

5

74

7%

31%

55%

7%

100%

aThis table and corresponding analyses include 73% of the codeable data (i.e., 1,143 out

of 1,562 codes), since in 27% of the segments students were engaged in off-task behavior.

2 Percentages cited are the total moves made by students and teachers.

(less than 1%) and a few instances of recycling goals in working memory (less than

1%). Teachers provided little in the way of instruction to students for planning

(only 3% of total codes) and virtually no scaffolding of these SRL behaviors (less

than 1% of total codes). The majority of teacher moves in this category were to

instruct students on their goals (2%). Teachers did not often activate students’

prior knowledge (less than 1%) and there were only a few instances of teacher

instruction of student planning (1%).

Monitoring

Chi-square analyses demonstrated that students and teachers also engaged in

very few behaviors associated with category of monitoring, accounting for 11% of

all the codes. Under monitoring, students and teachers were coded for behaviors

related to expressing a Judgment of Learning (JOL), partner questioning (PQ),

monitoring their progress toward goals (MPTG), teacher questioning (TQ),

and expressing a feeling of knowing (FOK). Student moves associated with

monitoring were responsible for 7%; but the majority of student moves in this

category were partner questioning (2%) and teacher questioning (4%). There were

few instances of judgment of learning (1%), feeling of knowing (less than 1%),

and monitoring progress toward goals (less than 1%). Teacher instruction and

scaffolding of monitoring accounted for 4%, with teachers helping students

monitor their progress toward goals (2% each for TI and TS). Teachers did not

scaffold or instruct students with respect to questioning, judgments of learning,

or feelings of knowing.

Strategies

Chi -square analyses demonstrated that students and teachers engaged in many

behaviors associated with category of strategies, accounting for 78% of all the

codes. Under strategies, students and teachers were coded for behaviors related to

following directions (FD – students only), giving procedural directions (Pro –

teachers only), content evaluation (CE), searching the environment (S), and

selecting a new informational source (SNIS). Student moves associated with

strategies were responsible for 63% of total codes. The majority of student moves

in this category were following directions (32%). Students also engaged in

selecting new information sources (18%), searching (8%), and evaluating content

(6%). These observations were expected, as the jigsaw tasks required these

behaviors for completion. Teacher scaffolding of strategies accounted for 4% of

moves, including helping students evaluate content (2%) and understand

procedures (1%). Teacher instruction of strategies accounted for 11% of moves,

including using procedures (which accounted for 10% of all moves) and

instruction on content evaluation (which accounted for 1% of all moves). Teachers

provided no other strategy help or instruction.

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 229

Task Difficulty and Demands

Chi-square analyses demonstrated that students and teachers engaged in very

few behaviors associated with category of task difficulty and demands, accounting

for only 7% of all the codes. Under task difficulty and demands, students and

teachers were coded for behaviors related to time and effort planning (TEP) and

help-seeking behavior (HSB). Student moves associated with monitoring were

responsible for 5% of total codes. The majority of student moves in this category

were help-seeking behavior (5%), usually concerning procedural questions.

Students engaged in virtually no time and effort planning (less than 1%).

In summary, although the analyses found significant differences between the

students’ mean pretest and posttest scores and in the frequency of the students’ and

teachers’ use of SRL variables, the learning gains were quite small. The following

four qualitative descriptions provide valuable insight into how these students

collaborated and self-regulated their learning of ecological systems with the WQS.

The typical pattern for students who were on-task working together is demon-

strated in example #1 (see Table 2). These students were involved in selecting

variables, looking at the graphs, and determining an appropriate value by evalu-

ating the graphs. Students seemed to be merely following directions (e.g., in turn 4,

“What do we put that under—nitrogen, right?”). There is no indication that

they were thinking critically about the information they were writing down.

When teachers were present and students demonstrated that they were

lacking the necessary knowledge to effectively evaluate the content in the hyper-

media environment, the teacher typically provided only procedural instruction. In

example #2, the teacher initiated an interaction with two students by providing

procedural instruction (see Table 2). However, when the teacher presented cog-

nitive scaffolding in the form of evaluating the content (e.g., in turn 5, “So,

what is this again?”), it becomes evident that the students lacked the necessary

knowledge and/or skills to effectively use this scaffolding. Recognizing this

inadequacy, the teacher returned to providing procedural instruction.

Students often sought help from the teacher while completing the RiverWeb

task. Example #3 demonstrates a typical teacher-student interaction in which

the students sought help with carrying out a basic procedure for the task, even

though the steps for the task had been explained to the whole class at the beginning

of the session (see Table 2). The student response to such procedural instruction

was merely to follow directions (e.g., in turns 10, 12, and 17, the student either

chose variables or wrote as directed); there is no evidence that the students were

attempting to understand the concepts.

Example #4 illustrates various scaffolding and instructional techniques that

teachers provided in order to facilitate students’ learning (see Table 2). The teacher

responded to a student-initiated interaction and provided scaffolding for a relation-

ship seen between indicators in the hypermedia environment. In this example, the

teacher used both instructional and scaffolding strategies, including procedural

230 / AZEVEDO, WINTERS AND MOOS

help (e.g., in turn 4, “Put a w.”) as well as scaffolding the evaluation of the content

(e.g., in turn 4, “Do you see it increasing or decreasing?”). In addition, the degree

and type of instruction and scaffolding provided by the teachers usually reflected

the level of students’ understanding of the information presented in the hyper-

media environment. In this example, cognitive scaffolding, as opposed to the more

superficial procedural instruction, was provided once the student demonstrated

understanding through evaluation of content (e.g., in turn 3, “Okay, this is a

weak relationship.”).

DISCUSSION

Our results show that low-achieving students tend to benefit little from Web-

based hypermedia learning environments designed to foster their learning of

challenging science topics. Despite performance data that show significant results,

the process data including the time and frequency of students’ SRL and teacher’s

scaffolding and instructional moves show the use of a disproportionate amount

of time and number of SRL moves during learning. The process data provide

evidence that students’ poor performance on the pretest-posttest shift is due to

their failure to deploy key SRL processes and mechanisms which could have led

to significant shifts in conceptual understanding. In addition, the same data show

that the teacher rarely deployed scaffolding and instructional moves aimed at

fostering students’ self-regulated learning of ecology.

With regard to the first research question, the results of this study show that

students did learn about the effects of land use during the two-week period in

which they used the WQS collaboratively in the ecology class. Despite the

statistically significant results on their scores from pretest to posttest, the shift

in scores is rather small (7%) and there is ample room for increased learning of

ecology, since the mean on the posttest was relatively low (19%). We can conclude

that while these Web-based hypermedia learning environments can potentially

foster students’ learning, not all students have the self-regulatory skills necessary

to learn from these rich environments. Our results indicate that providing low-

achieving students with complex, open-ended CBLEs without the appropriate

embedded and contextual scaffolding (e.g., from the teacher, peer, tutor, etc.) will

lead to inferior shifts in learning. This finding is consistent with the majority

of studies on non-linear, random access hypermedia environments with flexible

access and a high degree of learner control (e.g., Azevedo, Cromley, & Seibert,

2004; Azevedo, Guthrie, & Seibert, 2004; Greene & Land, 2000; Jacobson

& Archodidou, 2000).

This finding is consistent with previous research which indicates that students

lack the strategies, self-monitoring, persistence, and other self-regulatory skills

required to take advantage of the vast amounts of rich content contained in

CBLEs such as Web-based hypermedia learning environments. We hypothesize

that in addition to not having the self-regulatory skills, these students lack prior

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 231

232 / AZEVEDO, WINTERS AND MOOS

Table 2. Qualitative Descriptions of Students’ and Teacher Interactions

during Learning with RiverWeb [Non-Verbal Behavior] (SRL Code)

Turn Transcript

Example 1:

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Example 2:

1

2

Student 2: [Selecting variables] (FD)

Student 1: Toxins. . . . Alrighty then, right now we can look for

pristine forests. Let me see something. That don’t work. That sucks.

What does this say— nitrogen concentration.

Student 2: [unintelligible] Um, average is .7. (CE)

Student 1: What do we put that under—nitrogen, right? 0.7 [writing

on sheet]? (FD)

Student 2: [Unintelligible] [writing on sheet] (FD)

Student 1: [Selecting variables] There is runoff. (FD) (SNIS)

Student 2: 1.7- (CE)

Student 1: No, it is 1—1.797 (CE)

Student 2: I mean 1.797

Student 1: .5. [writing on sheet] (FD)

Student 2: Change it—to what? [Selecting variables] Sediments . . .

2.—(FD) (SNIS) (CE)

Student 1: - .6 (CE)

Student 2: 2.6 [writing on sheet] (CE) (FD)

Student 1: Water temp. [Selecting variables] (FD) (SNIS)

Student 2: . . . . And its . . . . [looking at screen and writing on sheet]

(CE) (FD)

Student 1: Phosphorous . . . [Selecting variables] (FD)

Student 2: Phosphorous.

Teacher: Okay, so you are going to complete this. This first thing

that I wanted you to do is put your station number here . . . Okay,

so you are going to, you are going to fill out everything here so we

know this belongs to. Ah, this is lumbering forest. You are going to

look at average runoff. This is nitrogen. (TI-PRO) What station is

this? (TS-PRO) Wait a minute, don’t, don’t change it yet. You already

have the average nitrogen, okay. (TI-PRO). What’s the average?

[referring to information in hypermedia environment] (TS-CE)

Student 1: 0.7865 (CE)

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 233

Table 2. (Cont’d.)

Turn Transcript

3

4

5

6

7

8

9

10

11

Example 3:

1

2

3

4

5

6

7

8

9

10

11

Teacher: Remember, I said you can, you can round that to .79,

would be good. (TI-PRO)

Student 1: Okay.

Teacher: Okay. So what is this again? [referring to information in

hypermedia environment[ (TS-CE) Where do you find that on your

table? Okay, for what station? (TS-PRO)

Student 2: I have no, I don’t know how to do none of this or

understand the problem. (JOL)

Teacher: Uh, uh. Fill that, put that in there. You will continue and fill

out the chart. (TI-PRO)

Student 1: What is this? (TQ)

Teacher: Well, you have, you have to fill this out first. (TI-PRO)

Student 1: How do I find about the [unintelligible]? (TQ)

Teacher: Well, again, click on here and it gives you all of your

different indicators. (TI-PRO)

Student 2: [Hand is raised] I need to find out what station we are.

(HSB)

Student 1: I think we are 6 [unintelligible].

Student 2: Yeah, I think we are 5 or 6. [Keeps hand raised]

. . . . . . . . . . . . . . . . . . . . . .

Student 1: Hey [to teacher] . . .

Teacher: Yes?

Student 1: What station? (HSB)

Teacher: Um, 6. So you can type in 6. (TI-PRO)

Student 2: What are we looking for now? (TQ)

Teacher: [referring to table on back of packet] Can I rip this? Ok.

You are going to click on your land use area. Then you are going to

use this table and you are going complete this for your indicators.

You are station 6, so I want you to write “6” in here [points to place

on sheet]. (TI-PRO)

Student 2: [Writes on sheet as directed] (FD)

Teacher: And then “urban”. Ok, now, you are going to complete this

entire table, so the first thing you are going to do (TI-P) is click on this

rectangle where it says air temp.—click there and choose runoff.

(TI-PRO)

knowledge of the topic and also lack specific science skills such as rounding

numbers. We acknowledge that one of the limitations of this study is that we used a

one-group pretest-posttest pre-experimental design and combined it with process

data. However, this was a necessary first step prior to conducting experimental

manipulations given the infancy of this type of research and the complex nature of

students’ SRL with CBLEs in the classroom (Mayer, 2003).

With regard to the second and third research questions, our extensive process

data indicated that students and the teacher differed in the amount of time and the

frequency of use of self-regulatory behaviors during learning with the WQS.

These students tend to spend more time on and use more self-regulatory skills,

followed by teacher instructional moves and then teacher scaffolding. Also, the

students and teacher use strategies as a predominant SRL variable during learning

234 / AZEVEDO, WINTERS AND MOOS

Table 2. (Cont’d.)

Turn Transcript

12

13

14

15

16

17

Example 4:

1

2

3

4

5

6

Student 2: [Clicks on rectangle and chooses runoff]. . . (FD)

Teacher: Ok, and then down here, what is the next one? (TS-CE)

Student 1: Sediment. (CE) [selects sediment]

Student 2: Sediment.

Teacher: Uh huh. Now when you change display—[points to screen]

you have minimum value, maximum value, and I want you to write

down the average, and again, you can round up. (TI-PRO) So that

would be 21,939. You write that here and you are done. (TS-MPTG)

Student 2: [Writes down number on sheet] (FD)

Student 1: Sketch that? [Question directed towards teacher] (TQ)

Teacher: Uh, uh. (TI-PRO)

Student 1: Okay, this is a weak relationship. [referring to scatter plot

graph in hypermedia environment] (CE)

Teacher: Uh, uh. Put a w. (TI-PRO). Negative? Do you see it

increasing or decreasing? (TS-CE). Or, if you can’t if you can’t tell if

it is going up or down, then there’s no relationship. (TI-CE)

Student 1: I can’t tell. [referring to scatter plot graph in hypermedia

environment] (CE)

Teacher: Okay, then you put no relationship. Okay, then as you

continue, you guys discuss it amongst yourselves. Make sure you

both agree. (TI-PRO)

with the WQS. The second most often used SRL variable by students and teachers

is monitoring. The quantity (i.e., amount and frequency of SRL variables) and

quality of the strategy and monitoring processes used by students may explain

why students gained so little knowledge of ecology with the WQS. The verbal

protocols provide process data to indicate that they used these SRL processes,

and together with the product data show that the use of these processes led to

significant increases in their understanding of the science topic. Students regulated

their learning by using mostly ineffective strategies and metacognitive monitoring

and did very little planning. In contrast, the teacher’s instructional and scaffolding

moves involved using strategies to facilitate students’ learning. These findings

highlight the value of converging product and trace data—i.e., while the statistical

analyses revealed significant differences in the time and frequency data, the

explanation lies in the analysis of the quality of each of these SRL moves used

by both students and teachers. For example, despite spending a significant amount

of time regulating their learning, and spending a disproportionate amount of time

using these strategies, the process data reveal students were using low-level

strategies such as following directions and searching the WQS without a specific

goal. Furthermore, the relative amount of time spent by the teacher on instructing

and scaffolding moves was relatively less than students spent, and the quality

of those moves can also be considered low-level in response to the students

level of understanding.

These findings are consistent with previous research on using trace method-

ologies to examine SRL processes used by learners during learning (Azevedo,

Cromley, & Seibert, 2004; Azevedo, Guthrie, & Seibert, 2004; Winne & Hadwin,

1998; Winne & Stockley, 1998). They also highlight current theoretical and

methodological concerns among SRL researchers that more research is required to

understand the inter-relatedness and dynamics of SRL variables during learning-

cognitive, motivational/affective, behavioral, and contextual during the cyclical

and iterative phases of planning, monitoring, control, and reflection (e.g., Butler,

2002; Meyer & Turner, 2002; Perry, 2002; Pintrich, 2000). This is definitively the

case in the area of learning with Web-based hypermedia learning environments,

where research tends to focus on learning gains (see Shapiro & Niederhauser,

2004 for examples) without considering the complexities of students’ SRL and

the teachers’ scaffolding of SRL during extended learning with CBLEs.

Implications for the Design of the CBLEs as

MetaCognitive Tools to Enhance Students’ Learning

The results of our study have implications and pose several challenges for the

design of the WQS to foster low-achieving students’ self-regulated learning

of ecology in technology-rich dynamic science classrooms. One challenge for

designing a Web-based hypermedia environment is the role of scaffolding, which

is critical in fostering students’ conceptual understanding and enhancing SRL. The

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 235

first issue related to scaffolding is the system’s ability to detect that students need

assistance in regulating their learning. For example, how can the WQS detect that

a student or the pair is/are using ineffective strategies or need(s) assistance in

monitoring their learning? Second, what types and levels of scaffolding methods

should be designed for low-self-regulating learners?

According to our results, these students typically use ineffective learning

strategies, do not monitor their learning, fail to engage in help-seeking behavior,

and have difficulties in handling task difficulties and demands. So, how can we

“build” system components (e.g., a student model, instructional model, interface,

inference engine, etc.) to determine if: a) a learner, in the complex interactions

during individual and collaborative knowledge construction is having difficul-

ties with several phases and areas of SRL; and b) what effects will this deter-

mination have on the detection, monitoring, and fostering of learners’ overall

self-regulation? How can we make our SRL model “visible” to the learners and

flexible enough to allow learners to explore advanced topics related to the

ecological systems, including its content and structure? How can the CBLE adapt

and exhibit flexibility during learning?

Recent research on pedagogical agents (e.g., Baylor, 2003; White et al., 2000)

suggest that these agents could be used to facilitate students’ SRL by scaffolding

students’ planning (e.g., include a planning net with a hierarchy of sub-goals,

model the phases of science inquiry), monitoring (e.g., prompting student to

periodically relate their emerging understanding with their prior knowledge and

instructional goals), strategy use (e.g., embed video clips that model effective

strategies such as making inferences and hypothesizing, model basic science

skills such as rounding off numbers), and handling task difficulties and demands

(e.g., prompts to help the student monitor their time and effort planning, provide

a digital library of video clips that deal with FAQs related to basic science

skills, inquiry skills, etc.).

Another critical issue is the system’s ability to construct an adequate model

of the learner’s emerging understanding of the domain (i.e., a mental model). How

can we design ways of detecting, monitoring, and fostering shifts in learners’

mental models of ecological systems? Can we have students create concept maps

and/or drawings which can be used to dynamically assess their existing mental

model, which could interact with other system components? For example, what

kind of instructional decisions should be made in a case where the system has

determined that a student has a sophisticated mental model of the ecological

system, but has expressed low interest in the task, versus a student who has

constructed a less-sophisticated mental model has expressed a high level of

interest in the topic, yet lacks the ability to plan learning and is using ineffective

strategies (e.g., “blindly” searching the WQS to find the representations of infor-

mation needed to explore the issues of the current goal)? Would making the learner

construct an “external” visual representation of his/her emerging mental model

of the domain allow him/her to self-regulate? At another level, would this visual

236 / AZEVEDO, WINTERS AND MOOS

representation allow a peer to assist with regulating their learning? Would this

external representation allow learners to jointly regulate their co-construction of

the topic? Can this information provide the WQS with other “variables” with

which to make informed instructional decisions? Also, would these external

representations, which are visible and accessible, allow the learner (and perhaps

others such as peers, and teachers) to share, inspect, critique, modify, and assess

the learners’ understanding of a complex system?

These issues pose several challenges for the learning sciences community. First,

there is a greater need to draw on the strengths of the members of our research

community in dealing with the complexities of studying SRL in complex instruc-

tional settings and in designing and developing CBLEs to trace, detect, and foster

students’ SRL. Second, the role of teachers needs to be examined in terms of how

different instructional methods and scaffolding support students’ SRL, and their

role as external regulatory agents creating instructional settings that promote

students’ SRL (Corno & Randi, 1999; Meyer & Turner, 2002; Perry et al., 2002).

Third, theoretical issues related to SRL challenge Derry and Lajoie’s (1993) initial

characterization of three camps for using computers as cognitive tools—modelers,

middle-campers, and non-modelers. Lajoie (2000) has recently argued that we

should abandon these “camps” due to both the influx of computer technologies

made available to teachers and educators and to recent challenges made by SRL

researchers about the complexity of learning. Future research should explore the

role of SRL and teachers’ fostering of students’ understanding of complex and

challenging science topics in dynamic technology-enhanced classroom learning

environments (Azevedo, 2002).

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 237

AP

PE

ND

IX

Cla

sses,D

escri

ptio

ns,an

dE

xam

ple

so

fth

eV

ari

ab

les

Used

toC

od

eLearn

ers

’S

elf-R

eg

ula

tory

Beh

avio

ran

d

Co

-Reg

ula

ted

Beh

avio

r(b

ased

on

Azeved

o,C

rom

ley,&

Seib

ert

,2

00

4).

Vari

ab

leD

escri

ptio

na

Exam

ple

Pla

nn

ing

Pla

nn

ing

(P)

Go

als

(G)

Pri

or

Kn

ow

led

ge

Activatio

n(P

KA

)

Recycle

Go

alin

Wo

rkin

gM

em

ory

(RG

)

Ap

lan

invo

lves

co

ord

inatin

gth

esele

ctio

no

f

op

era

tors

.Its

exe

cu

tio

nin

vo

lves

makin

g

beh

avio

rco

nd

itio

nalo

nth

esta

teo

fth

ep

rob

lem

an

da

hie

rarc

hy

ofg

oals

an

dsu

b-g

oals

Co

nsis

teith

er

ofo

pera

tio

ns

that

are

po

ssib

le,

po

stp

on

ed

,o

rin

ten

ded

,o

ro

fsta

tes

that

are

exp

ecte

dto

be

ob

tain

ed

.G

oals

can

be

iden

tified

becau

se

they

have

no

refe

ren

ce

to

alr

ead

y-e

xistin

gsta

tes

Searc

hin

gm

em

ory

for

rele

van

tp

rio

rkn

ow

l-

ed

ge

eith

er

befo

reb

eg

inn

ing

perf

orm

an

ce

of

ata

sk

or

du

rin

gta

sk

perf

orm

an

ce

Resta

tin

gth

eg

oal(e

.g.,

qu

estio

no

rp

art

so

fa

qu

estio

n)

inw

ork

ing

mem

ory

(WM

)

Stu

de

nt:

“Next

we

are

go

ing

tolo

ok

at

that”

[refe

rrin

gto

new

data

en

try]

Tu

tor

Instr

ucti

on

:“W

hen

yo

uare

do

ne,yo

ucan

loo

kat

sed

imen

ts”

Tu

tor

Scaff

old

ing

:“I

have

giv

en

yo

uan

exam

ple

to

use

toco

mp

lete

this

pag

e.”

Tu

tor

Instr

ucti

on

:“Y

ou

will

be

wri

tin

gre

latio

nsh

ips.”

Tu

tor

Scaff

old

ing

:“W

hy

mig

ht

sed

imen

tsan

d

nitro

gen

go

tog

eth

er?

”

Tu

tor

Instr

ucti

on

:“Y

ou

need

tore

mem

ber

wh

at

we

did

du

rin

gth

eto

ur.

”

Stu

de

nt:

“It

says

wri

teth

ere

latio

nsh

ips.”

238 / AZEVEDO, WINTERS AND MOOS

Mo

nit

ori

ng

Ju

dg

men

to

fLearn

ing

(JO

L)

Part

ner

Qu

estio

nin

g

(PQ

)

Mo

nito

rP

rog

ress

To

ward

Go

als

(MP

TG

)

Teach

er

Qu

estio

nin

g

(TQ

)

Feelin

go

fK

no

win

g

(FO

K)

Str

ate

gy

Use

Fo

llow

Dir

ectio

ns

(FD

)

Pro

ced

ura

l(P

ro)

Learn

er

beco

mes

aw

are

that

they

do

n’t

kn

ow

or

un

ders

tan

devery

thin

gth

ey

read

Learn

er

dir

ects

learn

ing

or

pro

ced

ura

lb

ased

qu

estio

nto

ward

sco

llab

ora

tive

part

ner

Assessin

gw

heth

er

pre

vio

usly

-set

go

alh

as

been

met

Learn

er

dir

ects

learn

ing

or

pro

ced

ura

lb

ased

qu

estio

nto

ward

ste

ach

er

Learn

er

isaw

are

ofh

avin

gre

ad

so

meth

ing

in

the

past

an

dh

avin

gso

me

un

ders

tan

din

go

fit,

bu

tn

ot

bein

gab

leto

recall

ito

nd

em

an

d

Learn

er

isin

vo

lved

ino

ne

ofth

efo

llow

ing

learn

ing

-based

beh

avio

rs:

(1)

Read

ing

wo

rksh

eet,

(2)

en

teri

ng

data

,(3

)calc

ula

tin

g,

or

(4)

co

llab

ora

tin

gw

ith

part

ner

Can

be

no

n-v

erb

alb

eh

avi

or

Teach

er

pro

vid

es

scaffo

lsd

ing

/in

str

uctio

n

on

pro

ced

ure

ofle

arn

ing

task

Stu

de

nt:

[Refe

rrin

gto

teach

er

scaffo

ldin

go

fco

nte

nt

evalu

atio

n]

“Id

on

’tkn

ow

ho

wto

use

tho

se

gra

ph

s.”

Stu

de

nt:

[Dir

ecte

dto

ward

sp

art

ner]

“Ho

wd

ow

e

kn

ow

wh

at’s

air

tem

p?”

Stu

de

nt:

“We

are

do

ne.”

Tu

tor

Scaff

old

ing

:“A

reyo

ug

uys

fin

ish

ed

?”

Tu

tor

Instr

ucti

on

:“Y

ou

sh

ou

ldb

esta

rtin

go

nyo

ur

seco

nd

ch

art

.”

Stu

de

nt:

[Dir

ecte

dto

ward

ste

ach

er]

“Wh

at

are

we

su

pp

ose

tod

ow

ith

the

calc

ula

tor

part

?”

Stu

de

nt:

“Id

idre

ad

them

”[b

ut

do

es

no

tkn

ow

the

an

sw

er]

.

Stu

de

nt:

Typ

es

data

into

co

mp

ute

r,calc

ula

tes,w

rite

s

info

rmatio

no

nw

ork

sh

eet,

dis

cu

sses

task

ste

ps

with

part

ner.

Te

ach

er

Scaff

old

ing

:[R

efe

rrin

gto

wo

rksh

eet]

“Wh

ich

lan

du

se

isit

for?

”

Te

ach

er

Instr

ucti

on

:“Y

ou

on

lyn

eed

toavera

ge”

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 239

AP

PE

ND

IX(C

on

t’d

.)

Vari

ab

leD

escri

ptio

na

Exam

ple

Co

nte

nt

Evalu

atio

n

(CE

)

Searc

h(S

)

Sele

ctin

gN

ew

Info

rmatio

nal

So

urc

e(S

NIS

)

Mo

nito

rin

gco

nte

nt

rela

tive

tog

oals

Searc

hin

gth

eh

yp

erm

ed

iaen

vir

on

men

t

with

ou

tsp

ecifyin

ga

sp

ecific

pla

no

rg

oal

Can

be

no

n-v

erb

alb

eh

avi

or

Th

esele

ctio

nan

du

se

ofvari

ou

sco

gn

itiv

e

str

ate

gie

sfo

rm

em

ory

,le

arn

ing

,re

aso

nin

g,

pro

ble

mso

lvin

g,an

dth

inkin

g.M

ay

inclu

de

sele

ctin

ga

new

rep

resen

tatio

n,co

ord

inat-

ing

mu

ltip

lere

pre

sen

tatio

ns,etc

.

Can

be

no

n-v

erb

alb

eh

avio

r

Stu

de

nt:

[Refe

rrin

gto

scatt

erp

lot

inh

yp

erm

ed

ia]

“Th

at’s

po

sitiv

e;

that’s

str

on

g.”

Tu

tor

Scaff

old

ing

:[R

efe

rrin

gto

scatt

erp

lot

in

hyp

erm

ed

iaen

vir

on

men

t]“D

oyo

usee

itin

cre

asin

g

or

decre

asin

g?”

Tu

tor

Instr

ucti

on

:[R

efe

rrin

gto

scatt

erp

lot

in

hyp

erm

ed

iaen

vir

on

men

t]:

“Th

ink

ab

ou

tw

hy

the

str

on

go

nes

are

str

on

g.”

Stu

de

nt:

Learn

er

man

ipu

late

sh

yp

erm

ed

iaen

vir

on

-

men

tb

yscro

llin

geith

er

up

or

do

wn

Stu

de

nt:

[Learn

er

en

ters

new

data

]th

en

ad

van

ces

ton

ew

line

or

scatt

erp

lot

gra

ph

240 / AZEVEDO, WINTERS AND MOOS

Task

Dif

ficu

lty

an

d

De

man

ds

Tim

ean

dE

ffo

rt

Pla

nn

ing

(TE

P)

Help

Seekin

g

Beh

avio

r(H

SB

)

Oth

er

Un

co

deab

le(U

C)

Off-t

ask

(OT

)

Att

em

pts

toin

ten

tio

nally

co

ntr

olb

eh

avio

r

Learn

er

seeks

assis

tan

ce

reg

ard

ing

eith

er

the

ad

eq

uate

ness

ofth

eir

an

sw

er

or

their

instr

uctio

nalb

eh

avio

r

Can

be

no

n-v

erb

alb

eh

avi

or

Du

eto

vid

eo

cam

era

an

gle

an

d/o

rau

dio

,

beh

avio

ran

d/o

rverb

aliz

atio

no

fle

arn

er(

s)

can

no

tb

ere

liab

lyco

ded

usin

go

ne

ofth

e

SR

Lco

des

Learn

er

exh

ibits

beh

avio

rth

at

iscle

arl

yn

ot

with

inth

ele

arn

ing

task,in

clu

din

gta

lkin

g

off-t

op

icw

ith

co

llab

ora

tive

part

ner

an

d/o

r

peer.

Can

be

no

n-v

erb

alb

eh

avi

or

Stu

de

nt:

[To

part

ner]

“Wh

at

do

we

do

no

w?”

Te

ach

er

Scaff

old

ing

:[R

efe

rrin

gto

on

eo

fth

e

co

llab

ora

tive

learn

ers

]“A

reyo

uh

elp

ing

him

?”

Tu

tor

Instr

ucti

on

:“I

will

giv

eyo

ua

co

up

lem

ore

min

ute

s.”

Stu

de

nt:

Rais

es

han

dan

d/o

rverb

ally

so

licits

help

fro

mte

ach

er

or

researc

her.

aA

llco

des

refe

rto

wh

at

was

reco

rded

inth

ecla

ssro

om

an

dcap

ture

do

nth

eau

dio

an

dvid

eo

tap

es

(i.e

.,w

hat

stu

den

tsre

ad

,saw

,h

eard

,o

r

did

),in

clu

din

gw

hat

the

teach

er

initia

ted

an

dscaffo

lded

du

rin

gle

arn

ing

with

the

WQ

S.S

om

eco

des

inclu

de

no

n-v

erb

alb

eh

avio

r.

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 241

ACKNOWLEDGMENTS

The authors would like to thank Stacy Pritchett for constructing the lesson

plans, assisting the classroom teachers, and scoring the pretests and posttests; and

Susan Ragan and Mary Ellen Verona at the Maryland Virtual High School for

allowing us to use RiverWeb in this classroom study. We also thank Jennifer G.

Cromley for comments on an earlier version of this manuscript; Laura Smith,

Lucia Buie, and Jennifer G. Cromley for assisting in the data collection; and

the high school teachers (Wendell Hall and Philip Johnson) and their students

for participating in this study.

REFERENCES

Azevedo, R. (2002). Beyond intelligent tutoring systems: Computers as MetaCognitive

tools to enhance learning? Instructional Science, 30(1), 31-45.

Azevedo, R., Cromley, J. G., & Seibert, D. (2004). Does adaptive scaffolding facilitate

students’ ability to regulate their learning with hypermedia, Contemporary Educational

Psychology, 29, 344-370.

Azevedo, R. Cromley, J. G., Winters, F. I., Xu, L., & Iny, D. (2003). Is strategy instruction

effective in facilitating students’ ability to regulate their learning with hypermedia?

In U. Hoppe, F. Verdejo, & J. Kay (Eds.), Artificial intelligence in education: Shaping

the future of learning through intelligent technologies (pp. 193-200). Amsterdam:

IOS Press.

Azevedo, R., Guthrie, J. T., & Seibert, D. (2004). The role of self-regulated learning in

fostering students’ conceptual understanding of complex systems with hypermedia.

Journal of Educational Computing Research, 30, 87-111.

Azevedo, R., Verona, M. E., & Cromley, J. G. (2001). Fostering students’ collaborative

problem solving with RiverWeb. In J. D. Moore, C. L. Redfield, & W. L. Johnson

(Eds.), Artificial intelligence in education: AI-ED in the wired and wireless future

(pp. 166-175). Amsterdam: IOS Press.

Baylor, A. (2003). Evidence that multiple agents facilitate greater learning. In U. Hoppe,

F. Verdejo, & J. Kay (Eds.), Artificial intelligence in education: Shaping the

future of learning through intelligent technologies (pp. 377-379). Amsterdam:

IOS Press.

Biswas G., Schwartz, D., Bransford, J., & CTGV. (2001). Technology support for complex

problem solving: From SAD environments to AI. In K. Forbus & P. Feltovich (Eds.),

Smart machines in education. Cambridge Press, MA: MIT Press.

Boekaerts, M., Pintrich, P., & Zeidner, M. (2000), Handbook of self-regulation. San Diego,

CA: Academic Press.

Butler, D. (1998). The strategic content learning approach to promoting self-regulated

learning: A report of three studies. Journal of Educational Psychology, 90(4), 682-697.

Butler, D. (2002). Individualizing instruction in self-regulated learning. Theory Into

Practice, 41(2), 81-92.

Clark, D., & Linn, M. (2003). Designing for knowledge integration: The impact of

instructional time. Journal of the Learning Sciences, 12(4), 451-493.

242 / AZEVEDO, WINTERS AND MOOS

Corno, L., & Randi, J. (1999). A design theory for classroom instruction in self-

regulated learning? In C. Reigeluth (Ed.), Instructional-design theories and

models: A new paradigm of instructional theory (Vol. II; pp. 293-318). Mahwah, NJ:

Erlbaum.

Derry, S. J., & Lajoie, S. P. (1993). Computers as cognitive tools (1st ed.). Hillsdale, NJ:

Erlbaum.

Greene, B., & Land, S. (2000). A qualitative analysis of scaffolding use in a resource-

based learning environment involving the World Wide Web. Journal of Educational

Computing Research, 23(2), 151-179.

Hickey, D., Kindfield, A., Horowitz, P., & Christie, M. (2003). Integrating cur-

riculum, instruction, assessment, and evaluation in a technology-supported

genetics learning environment. American Educational Research Journal, 40(2),

495-538.

Jacobson, M., & Archodidou, A. (2000). The design of hypermedia tools for learning:

Fostering conceptual change and transfer of complex scientific knowledge. Journal

of the Learning Sciences, 9(2), 149-199.

Jonassen, D. (1996). Computers as mind tools for schools. Columbus, OH: Merrill.

Kozma, R., Chin, E., Russell, J., & Marx, N. (2000). The roles of representations and

tools in the chemistry laboratory and their implications for chemistry learning.

Journal of the Learning Sciences, 9(2), 105-144.

Lajoie, S. P. (2000). Computers as cognitive tools II: No more walls: Theory change,

paradigm shifts and their influence on the use of computers for instructional purposes.

Mahwah, NJ: Erlbaum.

Lajoie, S. P., & Azevedo, R. (2000). Cognitive tools for medical informatics. In S. P. Lajoie

(Ed.), Computers as cognitive tools II: No more walls: Theory change, paradigm shifts

and their influence on the use of computers for instructional purposes (pp. 247-271).

Mahwah, NJ: Erlbaum.

Lajoie, S. P., Guerrera, C., Munsie, S., & Lavigne, N. (2001). Constructing knowledge in

the context of BioWorld. Instructional Science, 29(2), 155-186.

Mayer, R. (2003). Learning environments: The case for evidence-based practice and

issue-driven research. Educational Psychology Review, 15(4), 359-366.

McCaslin, M., & Hickey, D. (2001). Self-regulated learning and academic achievement:

A Vygtoskian perspective. In B. Zimmerman & D. Schunk (Eds.), Self-regulated

learning and academic achievement: Theoretical perspectives (pp. 227-252). Mahwah,

NJ: Erlbaum.

Meyer, D. & Turner, J. (2002). Using instructional discourse analysis to study the scaf-

folding of student self-regulation. Educational Psychologist, 37(1), 17-25.

Paris, S. G., & Paris, A. H. (2001). Classroom applications of research on self-regulated

learning. Educational Psychologist, 36(2), 89-101.

Perry, N. (2002). Introduction: Using qualitative methods to enrich understandings of

self-regulated learning. Educational Psychologist, 37(1), 1-3.

Perry, N., VendeKamp, K., Mercer, L., & Nordby, C. (2002). Investigating teacher-

student interactions that foster self-regulated learning. Educational Psychologist,

37(1), 5-15.

Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In

M. Boekaerts, P. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation

(pp. 451-502). San Diego, CA: Academic Press.

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 243

Rasku-Puttonen, H., Etelapelto, A., Arvaja, M., & Hakkinen, P. (2003). Is successful scaf-

folding an illusion? Shifting patterns of responsibility and control in teacher-student

interaction during a long-term learning project. Instructional Science, 31, 377-393.

Reiser, B., Tabak, I., Sandoval, W., Smith, B., Steinmuller, F., & Leone, A. (2001).

BGuILE: Strategic and conceptual scaffolds for scientific inquiry in biology class-

rooms. In S. Carver & D. Klahr (Eds.), Cognition and instruction: Twenty-five years

of progress (pp. 263-306). Mahwah, NJ: Erlbaum.

Schunk, D. (2001). Social cognitive theory and self-regulated learning. In B. Zimmerman

& D. Schunk (Eds.), Self-regulated learning and academic achievement: Theoretical

perspectives (pp. 125-152). Mahwah, NJ: Erlbaum.

Shapiro, A., & Niederhauser, D. (2004). Learning from hypertext: Research issues and

findings. In D. H. Jonassen (Ed.), Handbook of research for education communications

and technology (2nd ed). Mahwah, NJ: Erlbaum.

Slotta, J. D., & Linn, M. C. (2000). The knowledge integration environment: Helping

students use the Internet effectively. In M. Jacobson & R. Kozma (Eds.), Learning

the sciences of the 21st century. Mahwah, NJ: LEA.

Suthers, D., & Hundhausen, C. (2003). An experimental study of the effects of repre-

sentational guidance on collaborative learning processes. Journal of the Learning

Sciences, 12(2), 183-218.

Vye, N., Schwartz, D., Bransford, J., Barron, B., Zech, L., & CTGV. (1998). SMART

environments that support monitoring, reflection, and revision. In D. Hacker,

J. Dunlosky, & A. Graesser (Eds.), Metacognition in educational theory and practice

(pp. 305-346). Mahwah, NJ: Erlbaum.

White, B. Y., Shimoda, T. A., & Frederiksen, J. R. (2000). Facilitating students’ inquiry

learning and metacognitive development through modifiable software advisers. In

S. P. Lajoie (Ed.), Computers as cognitive tools II: No more walls: Theory change,

paradigm shifts and their influence on the use of computers for instructional purposes

(pp. 97-132). Mahwah, NJ: Erlbaum.

Williams, M. (1996). Learner control and instructional technologies. In D. Jonassen (Ed.),

Handbook of research on educational communications and technology (pp. 957-983).

New York: Scholastic.

Winne, P. H. (2001). Self-regulated learning viewed from models of information proc-

essing. In B. Zimmerman & D. Schunk (Eds.), Self-regulated learning and academic

achievement: Theoretical perspectives (pp. 153-189). Mahwah, NJ: Erlbaum.

Winne, P., & Hadwin, A. (1998). Studying as self-regulated learning. In D. J. Hacker,

J. Dunlosky, & A. Graesser (Eds.), Metacognition in educational theory and practice

(pp. 277-304). Hillsdale, NJ: Erlbaum.

Winne, P. H., & Perry, N. E. (2000). Measuring self-regulated learning. In M. Boekaerts,

P. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 531-566).

San Diego, CA: Academic Press.

Winne, P., & Stockley, D. (1998). Computing technologies as sites for developing

self-regulated learning. In D. Schunk & B. Zimmerman (Eds.), Self-regulated learning:

From teaching to self-reflective practice (pp. 106-136). New York: Guilford.

Zimmerman, B. (2001). Theories of self-regulated learning and academic achievement:

An overview and analysis. In B. Zimmerman & D. Schunk (Eds.), Self-regulated

learning and academic achievement: Theoretical perspectives (pp. 1-37). Mahwah,

NJ: Erlbaum.

244 / AZEVEDO, WINTERS AND MOOS

Zimmerman, B., & Schunk, D. (Eds.). (2001). Self-regulated learning and academic

achievement (2nd ed.). Mahwah, NJ: Erlbaum.

Direct reprint requests to:

Roger Azevedo, Ph.D.

University of Maryland

Dept. of Human Development

3304 Benjamin Bldg., 3304E

College Park, MD 20742

e mail: razevedo@umd.edu

SRL AND COLLABORATIVE LEARNING WITH A CBLE / 245