Do students’ self-perceptions influence their cognitive abilities, and how do we know? In this quasi-experimental study, we examine possible answers to this question. Some of the concepts relate to early work by Bandura and models that followed on how self-efficacy influences learning performance.
In the session's discussion about self-perceptions, many expressed a belief that self-perceptions influence learning, but only one had gathered statistics to confirm the belief. Several assumed that low self-perception would produce low performance, but after our session, reported that their thinking had now shifted; indeed, in some contexts low self-perception may lead to less persistence, but in others such as the one we provided, it may lead to greater persistence. We also had a poster session on this same research, and discussion was fairly nonstop during this time.
We looked at self-confidence before and after instruction as one component of self-perception. A confidence survey was created for each module of a course in which students rated their confidence in performing discrete tasks and objectives.
In 2007 and 2008, pre-instruction confidence ratings were persistently low for all units and tasks in the course while post-instruction confidence ratings were persistently high. Use of the confidence rating system seems to improve learning as measured by grades, when compared to deliveries of the same course that did not use such a system (2006).
In 2007 and 2008 in courses that used the confidence rating system, students reported that the confidence rating process was of high value, and this seems to be associated with all grades given in the course. We think that these ratings suggest higher student engagement with the learning tasks.
In 2008, a pretest was added as a measure of pre-knowledge, and scores were persistently low (no score above 8). In this course (multiple sections), the pretest combined with the confidence rating system seems to have produced dramatic improvements in grades.
Anecdotally, students reported to the instructor that they were informed of what to study in the confidence surveys, that some would copy the surveys and expand them to use them as study guides. The combination of the pretest with the confidence rating provided them with greater clarity about expectations.
The next stage of work for this study will be systematic statistical analysis of the data. We invited advice from anyone in the group who could recommend statistical analyses that would fit with our context. One person took detailed notes about our project and will be sending us follow up recommendations about how to proceed with more rigorous statistical analysis.
Presenters: Don Slater, Georgia Southern University & Ludwika Goodson, Embry-Riddle Aeronautical University
In the session's discussion about self-perceptions, many expressed a belief that self-perceptions influence learning, but only one had gathered statistics to confirm the belief. Several assumed that low self-perception would produce low performance, but after our session, reported that their thinking had now shifted; indeed, in some contexts low self-perception may lead to less persistence, but in others such as the one we provided, it may lead to greater persistence. We also had a poster session on this same research, and discussion was fairly nonstop during this time.
We looked at self-confidence before and after instruction as one component of self-perception. A confidence survey was created for each module of a course in which students rated their confidence in performing discrete tasks and objectives.
In 2007 and 2008, pre-instruction confidence ratings were persistently low for all units and tasks in the course while post-instruction confidence ratings were persistently high. Use of the confidence rating system seems to improve learning as measured by grades, when compared to deliveries of the same course that did not use such a system (2006).
In 2007 and 2008 in courses that used the confidence rating system, students reported that the confidence rating process was of high value, and this seems to be associated with all grades given in the course. We think that these ratings suggest higher student engagement with the learning tasks.
In 2008, a pretest was added as a measure of pre-knowledge, and scores were persistently low (no score above 8). In this course (multiple sections), the pretest combined with the confidence rating system seems to have produced dramatic improvements in grades.
Anecdotally, students reported to the instructor that they were informed of what to study in the confidence surveys, that some would copy the surveys and expand them to use them as study guides. The combination of the pretest with the confidence rating provided them with greater clarity about expectations.
The next stage of work for this study will be systematic statistical analysis of the data. We invited advice from anyone in the group who could recommend statistical analyses that would fit with our context. One person took detailed notes about our project and will be sending us follow up recommendations about how to proceed with more rigorous statistical analysis.
Presenters: Don Slater, Georgia Southern University & Ludwika Goodson, Embry-Riddle Aeronautical University
No comments:
Post a Comment