Contact: Bruce Howard or Steven Mcgee, 316 Washington Avenue, Wheeling, WV 26003, (304) 243-2388
Contact: Bruce Howard or Steven Mcgee, 316 Washington Avenue, Wheeling, WV 26003, (304) 243-2388
Contact: Bruce Howard or Steven Mcgee, 316 Washington Avenue, Wheeling, WV 26003, (304) 243-2388
Abstract
This instrument development study had two goals. First, we had the pragmatic goal of
developing a measurement tool that would be valid for research and useful for assessment and
intervention in classrooms. Second, we sought to develop theory in the area of self-regulated
learning. To accomplish these goals, we began with theory-driven ideas and existing
instruments, and combined these with constructs that emerged from exploratory factor analysis.
Our results indicate that metacognitive awareness and regulatory skills in a problem-solving
context are comprised of five independent factors: Knowledge of Cognition, Objectivity,
Problem Representation, Subtask Monitoring, and Evaluation.
This research had two phases. In the first phase, we reviewed current techniques for
measuring variables related to metacognition and self-regulation and decided on a self-report
measure— mainly for pragmatic reasons. Next, data was collected using two existing
inventories related to metacognition and problem solving. After eliminating items based on
reliability analyses, remaining items were factor-analyzed resulting in five factors. In the second
phase, a new inventory was developed focusing on these five factors. Again, reliability analyses
were conducted, and the remaining items were factor-analyzed.
Our results have important implications for education. They indicate that there are five
particular metacognitive and self-regulatory constructs relevant to problem solving. Being able
to identify and delineate these constructs further should allow our educational research and
teacher professional development teams to begin providing teachers with a set of tools and
training resources to help them promote student self-regulation in their classrooms. These
resources would be important to teachers who are concerned not only about what students learn
but also about how they learn it.
Further, our analyses indicate that the constructs measured by the IMSR are independent,
and therefore a student may show preferences or “styles” of metacognitive strengths and
weaknesses that depend upon his or her unique combination of constructs. If these “styles” can
be further understood and delineated, it might be possible to train students to habitually use
particular regulatory behaviors.
1 1
Contact: Bruce Howard or Steven McGee, 316 Washington Avenue, Wheeling, WV 26003, (304) 243-2388
AERA 2000 2
Background
Method
We began with the pragmatic goal of developing an instrument that would further our
research and could also be used extensively in classrooms across the country to help teachers
identify students’ self-regulatory strengths and weaknesses. It was not our intention to replicate
the work of our predecessors in this area. Instead, we wanted to develop an easy-to-use self-
report inventory for use with 12–18-year-olds that focused more specifically on metacognitive
awareness and regulatory skills for solving mathematical and scientific problems.
Phase One
In Phase One the objectives were: (1) to confirm the existence of factors related to
knowledge of cognition and regulation of cognition in the context of problem solving, and (2) to
extend our understanding of regulatory skills related to planning, monitoring, and evaluating. To
this end, we began with two existing public domain inventories, the Jr.MAI, version B (discussed
AERA 2000 3
Materials
Jr.MAI form 2 (Version B)
The Junior Metacognitive Awareness Inventory was developed by Dennison, Murphy,
Howard, & Hill (1996). In version B, for grades 7-9, students are instructed to read 18 items,
and for each item circle the answer that best described the way they are when doing schoolwork
or homework (1=never, 2=seldom/rarely, 3=sometimes, 4=often/frequently, 5=always).
Results
Item Analysis
Both instruments were treated as a single inventory for our purposes. Item means ranged
from 2.3 to 4.5, and standard deviations ranged from .71 to 1.1. The standardized coefficient
alpha was .8569. Upon eliminating items 11, 17, 18, and 20 from the HISP, the reliability rose to
alpha = .8911.
Factor Analysis
We conducted an exploratory factor analysis with the expectation that we would see four
factors: a knowledge of cognition factor from particular Jr.MAI items, and three regulation of
cognition factors (planning, monitoring and evaluation) from both Jr.MAI items and the
remaining HISP items (36 total items). For the purposes of extracting the most independent
constructs, we utilized a principle components extraction method using varimax factor rotation.
Initial results revealed an eight-factor solution with eigenvalues 1.043 and above, which
accounted for 55.2% of the sample variance. Visual inspection showed that factors 6, 7, and 8
were comprised of only a total of four items that loaded .35 and above. A content analysis of
these four items showed that they measured particular learning strategies (e.g., “I draw pictures
AERA 2000 4
Phase Two
In Phase Two, our goal was to create a new inventory specific to metacognitive
awareness and regulatory skills in the context of problem solving. To this end, we examined the
23 remaining items from the original two inventories and revised or rewrote them to increase
reliability, and wrote additional items to clearly demonstrate the existence of the five factors that
had emerged in Phase One.
Procedures/ Participants
We recruited 829 students from schools across the country in grades 6-12 to complete the
revised inventory (51% male, 49% female). The ethnic breakdown included 85.2% Caucasian,
7.5% Hispanic, 5% African American, 1.9% Asian American, and 0.8% Pacific Islander.
We titled the new instrument the Inventory of Metacognitive Self- Regulation (IMSR).
The IMSR included 37 items with a five-point Likert scale. For each of the 37 items, students
were instructed to circle the answer that best described the way they are when doing schoolwork
or homework (1=never, 2=seldom/rarely, 3=sometimes, 4=often/frequently, 5=always). The
items are shown in Table 1.
Results
The overall inventory demonstrated a reliability of alpha=.935. We conducted an
exploratory principle components factor analysis using a varimax rotation. The resulting
solution revealed five factors with eigenvalues over 1.12, which accounted for 51.6% of the
variance. Reliability for each factor ranged from alpha=.720 to alpha= .867. Table 1 shows the
factors, their descriptions, and the factor weights above .40. In addition, Table 1 shows three
items (asterisked) that weighed only moderately across several factors, or weighed heavily on
factors different than those hypothesized. For future research we would recommend removing or
revising these three items.
Discussion
We found Knowledge of Cognition to be an important factor in this instrument as it was
in the Jr.MAI and the adult MAI (Dennison et. al., 1996; Schraw & Dennison, 1994). The fact
that Knowledge of Cognition stands out so clearly and independently indicates that more work
should be done in understanding this construct. Prior research on metacognition and self-
AERA 2000 5
regulation has also discussed the importance of constructs similar to Monitoring and Evaluation,
but very little research has connected the factors of Objectivity and Problem Representation to
self-regulation.
Our results have important implications for education. They indicate that there are five
particular metacognitive and self-regulatory constructs relevant to problem solving. Being able
to identify and delineate these constructs further should allow our educational research and
teacher professional development teams to begin providing teachers with a set of tools and
training resources to help them promote student self-regulation in their classrooms. These
resources would be important to teachers who are concerned not only about what students learn
but also about how they learn it. Our analyses indicate that the constructs measured by the IMSR
are independent, and therefore a student may show preferences or “styles” of metacognitive
strengths and weaknesses depending upon his or her unique combination of constructs. If these
“styles” can be further understood and delineated, it might be possible to train students to
habitually use particular regulatory behaviors.
References
Bruning, R. H., Schraw, G. J., & Ronning, R. R. (1995). Cognitive psychology &
instruction. Englewood Cliffs, NJ: Merril.
Brown, A. L. (1978). Knowing when, where, and how to remember: A problem of
metacognition. Advances in Instructional Psychology, 1, 77-165.
Brown, A. L. (1980). Metacognitive development and reading. In R.J. Spiro, B.C. Bruce,
& W. F. Brewer (Eds.), Theoretical issues in reading comprehension (pp. 458-482). Hillsdale,
NJ: Lawrence Erlbaum.
Brown, A. L. (1987). Metacognition, executive control, self-regulation,, and other more
mysterious mechanisms. In F. Weinert & R. Kluwe (Eds.), Handbook of child psychology: Vol.
3. Cogntive development (pp. 263-340). New York: John Wiley.
Dennison, R. S., Krawchuk, C. M., Howard, B. C., & Hill, L. (1996). The development
of a children's self-report measure of metacognition. Paper presented at the annual meeting of the
American Educational Research Association, New York.
Fortunato, I., Hecht, D., Tittle, C. K., & Alvarez, L. (1991). Metacognition and problem
solving. Arithmetic Teacher, 38(4), 38-40.
Palincsar, A. M., & Brown, D. A. (1987). Enhancing instructional time through
attention to metacognition. Journal of Learning Disabilities, 20(2), 66-75.
Schraw , G., & Dennison (1994). Assessing metacognitive awareness. Contemporary
Educational Psychology, 19, 460-475.
Schwartz, N. H., Anderson, C., Howard, B., Hong, N., & McGee, S. (1998). The
Influence of configural knowledge on children's problem solving performance in a hypermedia
environment. Paper presented at the annual meeting of the American Educational Research
Association, San Diego.
Zimmerman, B. (1989). A social cognitive view of self-regulated academic learning.
Journal of Educational Psychology, 81, 329-339.
AERA 2000 6
Table 1
The Five Factors, Their Descriptions and Factor Weights Above .40
Subtask Monitoring: Breaking the problem down into subtasks and monitoring the
choice of learning strategies and completion of each subtask
I use different learning strategies depending on the problem. 0.6595
I use different ways of learning depending on the problem. 0.5746
I try to break down the problem to just the necessary information. 0.5530
I identify all the important parts of the problem. 0.5335
I try to eliminate information in the problem that I don’t need. 0.5025
I pick out the steps I need to do this problem. 0.4650
I think about what information I need to solve this problem. 0.4466
*I think about all the steps as I work the problem. 0.3871 0.3868 0.3110 0.3566 0.1446
*asterisked items weighed only moderately across several factors, or weighed heavily on
factors different than those hypothesized.
How do You Solve Problems?
Please read the following sentences and circle the answer that best describes the way you are when
you are trying to solve a problem. Think about a problem that you might see in a science or math
class.
• Think about when you have to solve a hard problem. What do you do before you start?
• What do you do while you work the problem?
• What do you do after you finish working the problem?
There are no right answers--please describe yourself as you are, not how you want to be or think you
ought to be. Your teacher will not grade this.
2. I think of several ways to solve a problem and then choose the best one. A B C D E
10. I think about how well I am learning when I work a difficult problem. A B C D E
13. I read the problem over and over until I understand it.. A B C D E
16. When it comes to learning, I can make myself learn when I need to. A B C D E
Never Seldom/ Rarely Sometimes Often/ Frequently Always
A B C D E
17. I ask myself how well I am doing while I am learning something new. A B C D E
18. I check my work all the way through the problem. A B C D E
25. When I am done with my schoolwork, I ask myself if I learned what I wanted to A B C D E
learn.
28. I try to break down the problem to just the necessary information. A B C D E
The Inventory of Metacognitive Self-Regulation ©. Copyright 1998-2000 by Wheeling Jesuit University/ Center for
Educational Technologies. All rights reserved.
First presented in Howard, B. C., McGee, S., Shia, R., & Hong, N. S. (2000, April). Metacognitive Self-Regulation and
Problem-Solving: Expanding the Theory Base Through Factor Analysis. Paper presented at the Annual Meeting of the
American Educational Research Association, New Orleans, LA.