Greene, J. A., Moos, D. C., Azevedo, R., Winters, F. I. (2008). Exploring differences between gifted and grade-level students’ use of self-regulatory learning processes with hypermedia. Computers & Education, 50, 1069-1083.
Introduction
Greene, Moos, Azevedo, and Winters are prolific researchers of students’ use of self-regulated learning (SRL) processes in hypermedia environments. Some of their other studies have explored scaffolding of SRL strategies, contributions of SRL to shifts in mental models, and relationships between goal structures and SRL strategy use. This study utilized the same materials (pre and posttests, software hypermedia environment, and think-aloud coding) as the authors’ previous studies, but explored the differences in SRL strategy use between grade-level and gifted students. Hypermedia environments are germane to the investigation of SRL strategy use because the environments encourage free exploration, placing great responsibility on the student for pursuing learning goals.
Previous research into the SRL characteristics of gifted learners revealed greater understanding of metacognitive principles in gifted learners, but survey data reported no difference in deployment of SRL strategies between gifted and grade-level learners. Greene, et al. believed the survey results were misleading because they represented a self-report of the beliefs of the participants about their own SRL strategy use, rather than an objective measure of strategy deployment gathered during the learning process. This study (Greene, et al., 2008) used coded think-aloud data to measure the frequency of SRL strategy use. The authors, refuting the results of the survey research, hypothesized that gifted students use key self-regulatory strategies more than the grade-level students.
The design and hypotheses of this study aligned with Pintrich’s theory of Self Regulated Learning, which followed four areas of SRL (cognition, motivation, context, and behavior), through a four-phase cycle of learning: planning, monitoring, controlling, and reflecting. One assumption of Pintrich’s SRL model stated that all learners had the potential to regulate their learning, but research has shown that all learners do not always effectively regulate their learning. Since SRL strategy use is associated with greater learning gains, the authors believed that some of the differences between the performances of gifted and grade level students could be explained by differences in the application of SRL strategy.
Methods
Participants were ninety-eight 7th graders with average or little knowledge of the circulatory system. Half of the students attended a gifted program for highly able middle school students. A commercial software application served as the hypermedia learning environment. Identical pre and posttests were administered in paper and pencil format. Participants were tested individually and asked to think aloud continuously while they learned.
Assessments of Pre- and post-test mental models, collected in essay form, assigned a score of low, intermediate, or high for the purpose of analysis. Scores for matching and labeling tasks and blood flow diagrams reported percentage of correct answers. The 65 hours of audio and video recorded think-alouds were transcribed and coded based on the same process developed by Azevedo and colleagues for previous SRL studies. Agreement between the two raters was established at .96 for mental model scoring and .98 for verbal protocols. Inconsistencies were resolved through discussion.
Results
The authors utilized ordinal regression to ascertain the impact of gifted classification and pretest score on students’ posttest scores on the mental model essay, using the high mental model posttest group as the reference. Both pretest score and gifted classification were significant predictors in the resulting model. A low pretest mental model score reduced the odds of a high posttest score by 96% (eb = .04), while participants in the gifted classification saw the odds of a high posttest score increase by 794% (eb = 7.94).
On measures of declarative knowledge, the students in the gifted classification performed better on posttests, after adjusting for pretest scores. 2 × 2 (classification vs. time, mixed design) repeated measures ANOVAs showed a significant interaction between group and time for matching task (F[1,96] = 46.92, MSE = 250.66, p = .001, η2 = .33), labeling task (F[1,96] = 23.41, MSE = 217.02, p = .023, η2 = .20), and flow diagram (F[1,96] = 10.97, MSE = 3.28, p = .001, η2 = .10).
Chi-square analyses of SRL strategy use found significant differences between participant classification groups for only five of the 33 SRL variables. Participants in the gifted classification used summarization, selecting new informational sources, and coordinating informational sources significantly more than the participants in the grade-level classification, who used finding location in environment and taking notes significantly more (α = .05).
Discussion
The results confirmed all hypotheses: gifted students outperformed grade-level students on all learning measures and differences between groups emerged in use of SRL processes. The authors believed that differences in performance may be partly attributable to the differences in SRL strategy deployment. They proposed that grade-level students who learn to use higher level SRL strategies might improve performance.
The correlational design of this study could not establish a causal relationship between performance and SRL strategy use, which limited the applicability of the findings. Limitations also included participant age, prior knowledge classification, and the characteristics of the chosen hypermedia environment, which was designed for reference use, rather than as a teaching tool for meeting specific learning goals.
The authors recommend the application of their findings to the scaffolding of hypermedia environments. They suggest embedding SRL prompts and feedback in hypermedia in order to encourage students’ use of effective strategies and discourage use of ineffective strategies. They envision a hypermedia environment that utilizes artificial intelligence technology to follow and coach students’ self-regulatory practices while they learn.
Critique
This study filled an important gap in the prior research by exploring differences between gifted and grad-level students’ use of SRL strategies. The article was clearly written, well-organized, and conveyed adequate content and context. The methods developed by the authors for measuring SRL strategy use are the centerpieces of their research, and show marked improvement over previous studies of similar topics.
The coding of think-aloud protocols developed by the authors over the course of several studies, takes a step towards objective measurement of SRL strategy deployment from the previous survey methods, but is still susceptible to erroneous report. Specifically, participants’ cognition may be influenced by the requirement to think aloud and their very awareness of their thought-stream could cause them to change behaviors and depart from their usual learning processes. This would not signal a weakness in the measurement of the participant’s use of SRL processes in the research setting, but would limit the relevance of the research results to classroom contexts where think-aloud is not encouraged.
Is it plausible that the difference in learning product between summarizing and note taking is great enough to influence test scores at all? Because previous research has shown that gifted students possess more knowledge of SRL strategies, it is very likely that these students also apply the strategies differently. Measuring the frequency of strategy deployment does not evaluate the quality of strategy application. Perhaps future research could explore depth of practice in the application of SRL strategies.
Conclusions
Greene et al. confirmed some of their suspicions about gifted students’ use of key SRL strategies, but did not see the differences they expected in the areas of planning, monitoring, managing task difficulty and demands, or interest. If this finding is confirmed in other research, it might influence the theoretical concepts of the SRL cycle by placing more importance on certain key phases and processes. Practical applications of these findings might include direct instruction for implementing key SRL strategies during the learning process.
