Improving Engagement and Motivation Through Feedback and Self-Reflection
The action research project analyzed how timely feedback and self-reflection can increase students’ engagement
and motivation for future learning. Research with upper elementary students supports the assumption that students benefit from receiving regular feedback (Philippakos & MacArthur, 2016; Sackstein, 2017). Feedback from both teachers and peers can help students learn from mistakes and understand new concepts (Sackstein, 2017). Self-assessment is a feedback strategy that can have long-term positive effects on students’ motivation for future learning (Van Loon & Roebers, 2017). When coupled with explicit instruction of metacognitive strategies, self-assessment can have an even greater effect on learning and students’ motivation (Grünke, Sperling & Burke, 2017; Sisquiarco, Rojas, & Abad, 2018).
In my fourth grade classroom, I do not give students daily opportunities to self-reflect on their learning in all
subjects, in part due to time constraints. Teachers in my district report that they do not feel like they have sufficient instructional time (TELL Colorado Survey, 2015) and do not have time for lesson closure and feedback (Crested Butte Elementary Staff, 2018). Through my research, I wanted to study effective methods for facilitating self-reflection and learn how these practice lead to higher motivation and engagement for learners. I aimed to offer my colleagues recommendations for feedback and self-reflection practices, and provide evidence of why it is important to spend instructional time on these practices.
The primary research question is How does self-reflection affect student engagement and motivation for future
learning? Specific questions to be addressed include:
How does the timing of self-evaluation and feedback on a student’s self-evaluation affect the accuracy of the self-evaluation?
How does time spent on self-reflection help students set personal goals for future learning?
How effective is it to take time away from instruction in order to practice self-reflection?
What teaching practices will allow teachers to achieve the level 5 practice of Quality Standard III: Element B where “students monitor and revise their learning goals based on feedback”? (Colorado State Model Performance Management System, 2019)
To address the research questions, I began with a literature review to discover what methods published
researchers were using to explore effective feedback, self-reflection, student engagement, and motivation. The literature review exposed limitations in studies that I hoped to address in my own research. The literature review revealed useful recommendations that helped me develop the student survey, follow up survey, and interview questions with colleagues. Once these tools were built, I decided to explore my research question by collecting qualitative data from students during mathematics instruction and through interviews with colleagues. To address one of the limitations in the study by Fyfe and Rittle-Johnson (2017), I wanted to examine how feedback with time for revision may affect students’ learning and motivation.
Study Location and Participants
All research took place at Crested Butte Elementary School in Crested Butte, Colorado and all subjects are students
or employees of Crested Butte Community School in the Gunnison Watershed School District. The 25 students recruited for the study are my current 4th grade students. I also asked four colleagues to participate in this study. I obtained assent of all minor participants, signed consent from all parents/guardians, and signed consent for all adult participants. Students were invited to participate in the study through a class meeting and parent email. Colleagues were invited in person. All participation was voluntary with no incentives or rewards. Of the 25 students recruited, 19 students agreed to participate. All four colleagues agreed to participate in the study interview. One of the 19 student participants had an extended absence planned during the unit 5 assessment and associated student surveys. This student was excluded from this action research due to absences, which made a total of 18 student participants with complete data sets.
I collected five different types of physical data. I collected four observations of student engagement during
independent math practice in the classroom using Anita Archer’s (n.d.) “Continuous Observation of On-Task Behavior”. (Appendix A) I collected self-evaluation data using the Everyday Mathematics self-assessment rubrics from the unit 4 and unit 5 assessments (McGraw-Hill, 2018). I varied when students completed this self-evaluation either at the beginning (as the curriculum authors designed) or end of the unit assessment. (Appendix B) I collected proficiency data to evaluate the accuracy of the self-evaluation by scoring the assessments using the Everyday Mathematics answer key (McGraw-Hill, 2018).
I used two versions of pencil and paper surveys to measure students’ motivation for future learning. The short
response survey asked students to set learning goals based on their self-evaluation, performance on the unit assessment, and any feedback they received from their peers or teachers. I used a second variation of the pencil and paper survey, a follow up survey, which asked student how they were feeling about their academic performance based on the learning goals they set for themselves. (Appendix C) Finally, I collected data from four colleagues using interview questions (with notes taken in pencil during the interview) that focused on strategies for effective feedback, facilitation of student self-reflection, and the effects of feedback and self-reflection on student motivation. (Appendix D)
To analyze classroom engagement, I determined the percentage of intervals in which on-task behavior occurred
during four independent work time observations. To maximize participant confidentiality, I coded all data from the mathematics assessments and student surveys with randomly assigned student numbers. Next, I assessed student mathematical abilities based on the number of questions correct on the unit assessments. I matched which questions were associated with each skill and then compared my scoring of students’ skills with students’ self-assessments of the same mathematical skills. I created tables to easily compare the type of feedback received (self-correction or teacher feedback) with the accuracy of students’ self-assessment. (Appendix E) I repeated this procedure with the unit 5 data and then created another table to look for a correlation between when the students completed the self-assessment and the accuracy of the self-assessment. (Appendix F)
I studied students’ motivation for future learning based on responses to the survey and the follow up survey. I
noted if students had achieved their learning goal or, if not, what type of support they requested. I created a table to notice if there was any correlation between the accuracy of student’s self-assessments and their achievement of their learning goal. I also looked to see if there was any correlation between the type of feedback students received and their achievement of learning goals. (Appendix G) I reviewed students’ self-assessment of metacognitive strategies on the initial surveys and follow up surveys. I created a table to notice if there was any correlation between the type of feedback students received and students’ assessment of metacognitive strategies. (Appendix H) Finally, I color coded interview notes from the four interviews with colleagues, looking for any common themes. I noticed trends with effective feedback and common ways that teachers facilitate student self-reflection. I noted teacher observations about how student motivation is affected by feedback and self-reflection.
Data Presentation and Interpretations
Classroom observations of on-task behavior showed an overall increase in student engagement from the
beginning to the end of the study (Figure 1).
Figure 1. Student engagement during classroom observations. These observations were completed during independent math practice. At the end of each observation, I counted the number of intervals in which on-task and off-task behavior occurred to determine the percentage of on-task behaviors.
Data analysis showed overall student-self evaluation to be accurate on both the unit 4 and unit 5 assessments
(McGraw-Hill Education, 2018). On the unit 4 assessment the accuracy of students’ self-assessment ranged from -1.000 to 0.429. One student scored themselves on average one point lower than their proficiency and two students scored themselves on average 0.429 of a point higher than their proficiency. The average accuracy for unit 4 was 0.040. On the unit 5 assessment, the accuracy of students’ self-assessment ranged from 0 to 0.500. No student rated themselves on average below their proficiency and two students rated themselves on average 0.500 of a point higher than their proficiency. Student self-assessments may be more accurate overall in this study because students were invited to revise their self-assessment after feedback (see examples in Appendix B).
Analysis revealed an inconsistent correlation between the timing of self-assessment and the accuracy of students’
self-evaluation of mathematical skills (Figure 2 and 3). On the unit 4 assessment, students who completed the self-assessment at the end of the assessment rated their skills on average as over one point higher than their proficiency showed. Students who completed the self-assessment at the beginning of the assessment rated their skills on average as over half a point lower than their proficiency showed. On the unit 5 assessments, there was no notable difference between the timing of the self-assessment and the accuracy of students’ self-evaluations of mathematical skills.
Timing of Self-Assessment
Figure 2. Shows timing of self-assessment and overall self-assessment accuracy on unit 4. E represents at the end of the unit assessment and B represents at the beginning of the unit assessment. The self-assessment (SA) of mathematical skills included three categories of proficiency: I can do this if I get help or look at an example (code =2), I can do this on my own. (code =3), I can do this on my own and explain how to do this. (code = 4) (McGraw-Hill, 2018). The codes for this SA rubric are related to our standardized grades. Accuracy of SA was determined by how many points above or below students self-assessed their skills in comparison with teacher scored skill proficiency.
Figure 3. Shows timing of self-assessment and student self-assessment accuracy on unit 5. E represents at the end of the unit assessment and B represents at the beginning of the unit assessment. Self-assessment accuracy is measure in points that students rated themselves on average above or below their mathematical proficiency scores.
Timing of Self-Assessment
Data analysis revealed a correlation between the type of feedback a student received and the accuracy of students’
self-evaluations on the unit 4 assessment with self-correction correlating with more accurate self-assessment of mathematical skills than teacher feedback. Students in the self-correction group had the opportunity to revise independently or with a peer. Most students worked with a peer. Figure 4 highlights that peer feedback (SC group) may be just as influential as teacher feedback with self-assessment. This finding matches the research by Sackstein (2017) about the positive value of peer feedback. There was no notable correlation between the type of feedback received and average self-assessment accuracy with the unit 5 assessment (Figure 4).
Figure 4. Correlation between type of feedback students received on each assessment and overall accuracy of self-assessment of mathematical skills.
Unit 4 Type of Feedback Recieved
Average of Overall SA Accuracy
Unit 5 Type of Feedback Recieved
Average of Overall SA Accuracy
Reviewing the data, I found no correlation between the type of feedback students received and their academic
progress toward self-selected learning goals. After the unit 4 assessment, of the students who received teacher feedback 3 achieved their learning goal by the follow up survey and 4 did not. Of the students who self-corrected their assessments, 5 achieved their learning goal by the follow up survey and 2 did not. After the unit 5 assessment, 5 students in both the teacher feedback and self-correction groups achieved their learning goals by the follow up survey and 2 did not. One limitation of this analysis may be that the follow up survey was given too quickly after the learning goals were set and students overall needed more mathematics lessons and independent practice time to feel like they had achieved their learning goals.
I also found no correlation between the type of feedback a student received and their metacognitive strategies
(Figure 5 and 6). Students who received an answer key and time to work alone or with a partner to revise any mistakes on the unit assessment did not claim to use metacognitive strategies any more often than student who worked with the teacher to revise mistakes on the unit assessment. Overall scoring of metacognitive skills did increase from 2.94 after unit 4 to 3.17 on unit 5 with 2=never, 3=sometimes and 4=always.
Figure 5. The type of feedback received on unit 4 assessment and students’ self-assessment of metacognitive skills. Type of Feedback received was either teacher feedback (T) or self-correction (SC). Students’ self-assessment of metacognitive strategies (MS) is shown for the student survey (SS) following each unit assessment and the follow up survey (FUS) approximately two weeks after the initial survey. On all surveys, metacognitive strategy one (MS1) was “After a math lesson, I take time to assess my strengths and areas for growth”. On all surveys, metacognitive strategy two (MS2) was “Before solving a challenging problem, I make a plan.” Self-assessment of metacognitive strategies one and two (MS1 and MS2) were rated by three categories of: Never (code =2), Sometimes (code =3), Always (code = 4).
Figure 6. The type of feedback received on unit 5 assessment and students’ self-assessment of metacognitive skills.
Figure 7 and 8 explore whether students had set a new learning goal (NGS) or identified the support they
needed to achieve their learning goal, either wanting lessons/re-teaching (WR) or wanting independent practice time (WIP).
Figure 7. New goal set or type of support requested after unit 4. NGS = new goal set (originally leaning goal was achieved), WIP = want more independent practice time, and WR = want re-teach or lesson of specific skill.
Figure 8. New goal set or type of support requested after unit 5. NGS = new goal set (originally leaning goal was achieved), WIP = want more independent practice time, and WR = want re-teach or lesson of specific skill.
Of interest in these two pie charts is that students wanted independent practice time more than specific lessons on
the skill that they had identified as their learning goal. The challenge with facilitating this independent practice of student selected learning goals is the wide variety of skills that students felt proud of post unit assessment and the variety of skills that students chose as mathematical learning goals (Figure 9 and 10). After the unit 4 assessment, 18 students identified 9 unique skills that they were proud of and 7 different skills as their learning goals. After the unit 5 assessment, the same students identified 9 different skills they were proud of and 9 different skills as their learning goals.
Figure 9. Number of students versus mathematical skill identified in learning goal after unit 4 assessment.
Figure 10. Number of students versus mathematical skill identified in learning goal after unit 4 assessment.
Teacher interviews highlighted the value of open response lessons with all four teachers identifying open response
activities with open questioning as times where teachers and peers can provide valuable feedback and allow time for revision (McGraw-Hill, 2018). A common theme that four out four teachers emphasized was giving students time after feedback to revise, two teachers noted that time to apply feedback is a limiting factor in daily instruction. Two teachers mentioned that small groups allow them to provide individual feedback during math.
Two teachers highlighted the importance of positive feedback and focusing on what student can do through verbal
praise. These teachers observed that positive feedback may lead to higher motivation because it gives students confidence in what they are doing well and makes them feel like they can improve.
Teachers highlighted that a key component of self-reflection is regularly reviewing student goals. A teacher stated,
“Goals need to be referred to more than one time to be effective”. One effective strategy was to write student goals in a place where they are always visible to students. Two teachers mentioned that they have students write their own goals. One teacher explained that reflection about whether an answer makes sense or if there is another possible solution affects motivation because students are motivated to find new challenge through innovative solutions. Three teachers discussed failure and how important it is to celebrate failure as part of the learning process and then teach students new strategies for success.
Archer, A. (n.d.). Continuous interval observation of on-task behavior. Observation forms peer coaches. Retrieved Sunday,
October 7, 2018 from http://www.corelearn.com/files/Archer/Observation%20Forms%20Peer%20Coaches.docx
Colorado State Model Performance Management System (2018, April 25). Evaluator assessment rubric for Brynn O’Connell.
Retrieved September 8, 2018 from https://copms.randasolutions.com/Assessment
Crested Butte Elementary Staff. Crested Butte 4th grade teaching team, and Crested Butte literacy coach. (August 2018).
Fyfe, E. R., & Rittle-Johnson, B. (2016). Mathematics practice without feedback: A desirable difficulty in a classroom
setting. Instructional Science, 45(2), 177-194. doi:10.1007/s11251-016-9401-1
Grünke, M., Sperling, M., & Burke, M. D. (2017). The impact of explicit timing, immediate feedback, and positive
reinforcement on the writing outcomes of academically and behaviorally struggling fifth-grade students. Insights into Learning Disabilities, 14(2), 135-153.
Logan, J. (2014). School leadership through action research. Upper Saddle River, NJ: Pearson Education, Inc.
McGraw-Hill Education. (2018). Everyday mathematics grade 4. Teacher's edition. New York City, NY, USA: McGraw-Hill.
Philippakos, Z. A., & Macarthur, C. A. (2016). The effects of giving feedback on the persuasive writing of fourth- and fifth-
grade students. Reading Research Quarterly, 51(4), 419-433. doi:10.1002/rrq.149
Sackstein, S. (2017). Peer feedback in the classroom: Empowering students to be the experts. Alexandria, VA, USA: ASCD.
Sisquiarco, A., Rojas, S. S., & Abad, J. V. (2018). Influence of strategies-based feedback in students’ oral performance.
HOW, 25(1), 93-113. doi:10.19183/how.25.1.402
Teaching, Empowering, Leading and Learning (TELL) Colorado Survery (2015). Gunnison Watershed Re1J. Retrieved
September 8, 2018, from http:www2.cde.state.co.us/tell/historicaldata.htm
Van Loon, M. H., & Roebers, C. M. (2017). Effects of feedback on self-evaluations and self-regulation in elementary school.
Applied Cognitive Psychology, 31(5), 508-519. doi:10.1002/acp.3347