HW Assignment: Measurement of self-regulated learning and motivation

Erin E. Peters

EDEP 654 – Spring 2006

Instruments designed to measure self-regulated learning and motivation give researchers information about the success of the aspects of the constructs assessed on the instrument. Instruments used to measure self-regulation depend on the theoretical view of the researcher. Some of the theoretical views of self-regulation include operant, phenomenological, social cognitive, volitional, Vygotskian, and cognitive constructivist. Each theoretical view would focus on different details in their measurement instrument. For example, Vygotskian self-regulation would focus on inner speech and operant self-regulation would focus on behavioral aspects in their measurement tools. Although each theoretical view has a different lens from which to examine self-regulation, there are common factors that can be measured regardless of the theory of origin. Zimmerman (1989) describes five common factors of self-regulation: motivation, processes to become self-aware, key processes used to attain goals, social and environmental factors, and the acquisition of self-regulation when learning. When instruments are designed to measure self-regulation, they include some if not all of the five identified factors.

Motivational theories are more numerous than self-regulation theories and each tend to focus on slightly different aspects of human cognition or behavior. Ford (1992) identifies thirty-two different motivational theories that have been used since the early 1900’s. Again, instruments designed for each motivational theory would measure slightly different aspects of human behavior or cognition. For example, Motivational Systems Theory, a comprehensive theory, would measure some aspect of personal goals, goal content, goal hierarchy, goal-setting processes, goal orientation, and intentions. Control Systems Theories of Human Motivation would measure system goals, goal hierarchies, and error sensitivity. Overall, most theories include some facet of goal setting, attribution, evaluation of performance, and a response to the evaluation. An instrument measuring motivation would consider some of all of these characteristics of motivation. It is unreasonable to expect that instruments would be able to measure all aspects equally well, so most instruments focus on one or several aspects deeply in order to develop a picture of the human functions involved in self-regulation and motivation.

The measure I selected for self-regulation is the Metacognitive Orientation Scale-Science (MOLE-S) developed by Gregory P. Thomas who is a professor at The Hong Kong Institute of Education. The instrument is designed with a social constructivist view in mind and considers that knowledge is not constructed in a vacuum, but is developed through interactions with the learning environment. Thomas (2003) argues that most measures in the science classroom regarding metacognition involved lengthy interviews and observations and that the development of a large-scale measure of metacognition in the classroom would be useful. Eight aspects of metacognition which were supported by the research literature were measured on the MOLES-S: (1) metacognitive demands, (2) teacher modeling and explanation, (3) student-student discourse, (4) student-teacher discourse, (5) student voice, (6) distributed control, (7) teacher encouragement and support, and (8) emotional support. The MOLES-S is a 67-item instrument that includes the eight aforementioned dimensions based on a Likert-scale. The initial instrument was administered to 1026 students within the 14-17 year old age group. At the time the instrument was administered, Hong Kong school had five bands of stratification for student ability and achievement. The instrument was administered to equal numbers of students among each of the five groups. The initial instrument was refined used a Cronbach alpha coefficient analysis and was changed from eight scales to seven. The Teacher Modeling and Explanation scale overlapped with the Metacognitive Demands scale and lead to the deletion of the Teacher Modeling and Explanation scale. The refined MOLE-S reported an Alpha reliability ranging from 0.72 to 0.87 for each of the seven scales and all of the scales showed to be statistically significant. The discriminant validity ranged from 0.34 to 0.49 for each scale. The data suggested that students have little control over classroom activities, that students are on average in terms of metacognitive ability, and do not tend to discuss the process of learning science with teachers.

The instrument that I chose for the motivation portion of this assignment is the Changes in Attitude about the Relevance of Science (CARS) Questionnaire by M. A. Siegel and M. A. Ranney from the University of California, Berkeley. The items regarding the relevance of science were pilot-tested with 188 ninth and tenth graders, analyzed using a Rasch modeling, and then equally distributed on three equivalent questionnaire forms. The three different 25-item questionnaires were developed from the field-test questionnaire of 69-items so that retest effect was reduced in the two groups. The items were also equally balanced in terms of difficulty in agreement or disagreement. Each of the forms was then used in conjunction with two different interventions, Science and Sustainability, and Convince Me. There were 28 tenth grade students in the Science and Sustainability group and 19 tenth grade students in the Convince Me group. The two classes scores on a standardized test (Terra Nova test) were not significantly different (p= .35) before the intervention. The questionnaire was administered as a pre-test, a post-test and a delayed post-test. The internal consistency of each scale was above .80. In the final analysis, both groups tended to agree more over time that science is relevant to their lives. The 3x2 (Time x Participant group) ANOVA showed no significant difference between groups and revealed a significant difference over time (F(1,18) = 17.98, p‹.000). The overall gain was equivalent for both classes, increasing from .45 to .46. Although this instrument showed to be both reliable and valid, I think it would be a stronger measure if the 25 items were categorized into scales rather than merely asking for general perceptions.