A case study of peer assessment at BSix Executive Summary The aim of this research was to examine the impact of using peer assessment to improve students’ writing in one AS-level lesson with eight students. Peer assessment has been used as an ongoing tool with this class for approximately five months. The main findings include: Seven of the eight students improved their final score following the use of peer assessment. The maximum score for the exercise was 14 marks. Two students improved their score by nine to ten points, three students by seven points, two by five points and one by one point. One student’s final score however dropped by five points. Seven of the eight students felt peer assessment had helped them improve (on a scale of 1-4, in which 1= no improvement and 4= much improvement) by an average of 2.3 to 3.8. Five of these seven students showed a considerable improvement in their final score. The most common advantage of peer assessment mentioned by four students was that peer assessment helped them understand how to write better to attract marks. In relation to disadvantages, students were concerned that their peers were not qualified to mark their papers. 1 Introduction This small piece of research was commissioned to explore the practice and impact of peer assessment based on the work of one teacher in the Humanities department. This will act as a starting point to BSix Brookehouse Sixth Form College’s commitment to increase the practice and quality of peer assessment within the Humanities department and throughout the college. According to Sebba et al (2008) peer-assessment involves students assessing each others’ work, through reflection on the goals and what it means to achieve them. Peer-assessment may take place in pairs or groups, where the aim may be as much the development of group processes as the promotion or judgement of individual learning. Peer-assessment has particular value in formative assessment since students ask of each other questions they may be inhibited from asking their teacher, and explain things to each other using familiar language. The peer assessment model takes three main skills into account: (1) defining assessment criteria: thinking about what is required and referring to the product or process; (2) judging the performance of a peer: reflecting upon and identifying the strengths and weaknesses in a peer's product and writing an assessment report; and (3) providing feedback for future learning: giving constructive feedback about the product of a peer (Sluijsmans, 2002). A systematic review of research evidence of the impact on students in secondary schools of self and peer assessment was conducted by Sebba et al (2008). They found that student attainment increased across a range of subject areas, students’ self esteem increased, students improved in learning to learn, specifically goal setting, clarifying objectives, taking responsibility for learning and increased confidence. Methodology This research was a pilot study, and took the form of a single case study that examined the impact of using peer assessment to improve students’ writing in one AS-level lesson. Peer assessment has been used as an ongoing learning tool with this class for approximately five months. This research project was a collaboration between the class teacher and the BSix researcher-in-residence. 2 The research was administered in the following way: Class activity Research method Writing exercise : 5 minutes time Questionnaire administered to students at end constrained individual activity (See of activity to measure their perceptions of appendix 1). their performance (see appendix 2). Post session: Class teacher to mark work using an objective comparative marking template (see appendix 3). Individual assessment of peer’s work using Observation (objective and subjective) of the mark scheme and recommended writing process by class teacher. frame template - annotate script, allocate marks (See appendix 4). Verbal feedback between peer’s Giving marks, commenting on script, making Observation (objective and subjective) of the process by class teacher. recommendations. Rewrite answer to original question Questionnaire administered to students at end of activity to measure their perceptions of the impact of peer assessment, and elicit their views on the benefits/challenges of peer assessment (see appendix 5). Post session: Class teacher to reflect on process using notes from observation. Researcher-in-residence to analyse questionnaire results. Class teacher to mark 2nd draft of student writing using an objective comparative marking template. Final discussion and analysis by class teacher and researcher-in-residence. Ethics Eight of the nine students in the class gave their informed consent to take part in the research. Student nine did not take part in the research. All participants have been given anonymity in the report writing. See appendix 6 for the letter given to students about the research. 3 Findings Student Pre score (max 14) Post score (max 14) Student perceptions Scores are based on the mean of five questions that were marked by students on a scale of 1-4 Questionnaire one on initial performance 1= not at all 4= a lot (see appendix 2) Questionnaire two on impact of peer assessment 1= no improvement 4= much improvement (see appendix 5) 1 3 13 Student 1 perceived their initial performance as 2.6. Overall student 1 felt that peer assessment had helped them improve by a mean of 3.8. Student 1’s overall scored increased by 10 points. 2 6 13 Student 2 perceived their initial performance 2.4. Overall, student 2 peer felt that peer assessment had only helped them improve by a mean of 1.4 Despite this perception, student 2’s scored increased by 7 points. 3 6 13 Student 3 perceived their initial performance as 3.2. Overall student 3 felt that peer assessment had helped them improve by a mean of 3.2. Student 3’s score increased by 7 points. 4 13 8 Student 4 perceived their initial performance as 3. Overall student 4 felt that peer assessment had helped them improve by a mean of 3. Student 4’s score however dropped by 5 points. 5 0 7 Student 5 perceived their initial performance as 2. Overall student 5 felt that peer assessment had helped them improve by a mean of 3. Student 5's score increased by 7 points. 6 3 4 Student 6 perceived their initial performance as 2. Overall student 6 felt that peer assessment had helped them improve by a mean of 2.3 Student 6's score increased by 1 point. 7 3 8 Student 7 perceived their initial performance as 1.4. Overall student 7 felt that peer assessment had helped them improve by a mean of 2.4. Student 7's score increased by 5 points. 8 1 10 Student 8 perceived their initial performance as 1.8. Overall student 8 felt that peer assessment had helped them improve by a mean of 3. Blank's score increased by 9 points 4 Discussion When we compare pre-peer assessment and post-peer assessment scores, we find that seven of the eight students improved their final score. The maximum score for the exercise was 14 marks. Two students improved their score by nine to ten points, three students by seven points, two by five points and one by one point. Although we see some substantial improvements, it should be noted that if students got one aspect of the question right, this gave them the advantage of getting other aspects of the question right. Of the eight students, one student’s final score however dropped by five points (initial score 13/14 to 8/14). This could be a result of various factors including poor peer assessment feedback, confusion or tiredness. Overall, seven of the eight students felt peer assessment had helped them improve (on a scale of 1-4, in which 1= no improvement and 4= much improvement) by an average of 2.3 to 3.8. There is a correlation between their perception and their post- peer assessment score in which five of the seven students showed a substantial improvement. As noted above, the final score of the one student (student four) did however drop. The other student (student seven) gave themselves an initial average score of 2, and felt that peer assessment had helped by 2.3. In this case, the final score only improved by one point. Although the remaining student (student two) did not think peer assessment had helped a great deal (1.4), he/she still produced a piece of writing that resulted in a substantial improvement in their final score. Students’ perceptions can be analysed in further detail. This can be done by comparing perception of the initial score with perceptions of the impact of peer assessment on improving the writing task.1 This analysis is based on eight students marking a total of 39 statements on the 1-4 scale.2 We find that in the 20 cases where students perceived their initial performance as 1 or 2, they felt that peer assessment had helped them improve, in many cases by 3 or 4. Even in the 12 cases where students scored a 3 for their perceived initial performance, they still found that peer assessment had helped them improve. In one case where perceived initial performance was given the maximum score of 4, peer assessment was still seen to help by 3. Thus the 1 2 This analysis is based on student perception and does not take into account final score outcomes. Each student marked five statements, with the exception of one student that left one statement blank. 5 practice of peer assessment appears to be a useful learning tool, even for students who feel they are performing well. Perception of initial performance (1= not at all 4= a lot) Perception of impact of peer assessment (1= no improvement 4= much improvement) 1 1 2 2 2 4 3 4 3 2 3 3 4 3 4 3 No. of occurrences 1 2 5 8 4 Total 20 3 9 Total 12 1 Total 1 There were four cases where students perceived their initial performance as 1 or 2 and then felt that peer assessment had not helped them improve. This was in relation to the following statements: peer assessment has helped me to structure my paragraph (noted by two students), peer assessment has helped me understand what I need to write (noted by one student) and peer assessment has helped me understand how to write to attract marks’ (noted by one student). It is not necessarily surprising to find that in two cases, where a student perceived their initial performance was 3, they did not find peer assessment helpful. Perception of initial performance (1= not at all 4= a lot) 1 2 3 Perception of impact of peer assessment (1= no improvement 4= much improvement) 1 1 1 No. of occurrences 2 2 2 6 Students’ perceptions of advantages and disadvantages of peer assessment Students were also asked to comment, in writing, on the advantages/benefits and disadvantages/challenges of peer assessment. The most common advantages mentioned by four students was that peer assessment helped them understand how to write better to attract marks. In relation to disadvantages, students were concerned that their peers were not qualified to mark their papers. Students’ feedback is recorded below. Advantages/benefits of peer assessment Good because you can obtain first hand understanding of the mark scheme and there is a sense of familiarity Shows you want you need to write and how to attract marks More likely to take on board You can get feedback on your work. It helped me improve on how to get more marks on the question. Helps you understand exactly how to attract marks See what other people have answered Helps you stay engaged in learning [and] how to improve work. Hear different opinions and it is more fun Disadvantages/challenges of peer assessment The students do not have the knowledge or experience to mark and explain improvements They tell you how to paragraph Not easy if peer doesn’t understand the question You might not agree with the mark they gave you You can not read [other students] writing Don’t understand how the answers relate to the questions, just stated with no explanation The peer maybe wrong or they don’t listen or take in what you have got to say 7 Limitations of study This was a small pilot study conducted with a limited number of students (total=8). It is possible that peer assessment was not the only explanation for improvements in students’ final writing scores. It is therefore important to highlight some of the other variables that could have had an impact on the research findings: Unrepresentative sample and no control group to measure impact of other variables. Confidence levels regarding knowledge and writing was low before they started the first writing exercise. The opportunity to repeat the writing exercise a second time could in itself have resulted in a better score. Students may have picked up the answer from reading their peer’s work, rather than through the ability to peer assess. Hawthorne effect: students knowing that they were part of a piece of research and therefore using the lunch period to review the exercise and find the right answer (although this should not necessarily be seen as negative). The study only measured the immediate impact of peer assessment. It did not measure the medium and long term impact i.e. if students were asked to re-write the paragraph the following day, there is no guarantee that their scores would remain at the higher end of the scale. The study should therefore be repeated with a control group to see if other variables are having an impact on the study. The study should be extended as part of an action research project to see if the results are replicable over a number of cycles. The findings should be shared with the students who took part in the research and should include discussions about the disadvantages they associate with peer assessment (see appendix 7 for summary of feedback for student participants). The findings should also be shared with other teachers in the college to encourage them to both practice and evaluate peer assessment. Conclusion: Teacher’s reflections on the process The explicit aim of this small piece of research was to test the effectiveness of peer assessment in developing academic literacy skills in A-level Sociology; the implicit aim was to establish irrefutable data which would get learners to embrace peer assessment as a learning strategy. To this extent, I believe that both the method and the outcome have achieved these aims. The use of my learners in the process has developed their appreciation of peer assessment as evidenced by their unquestioning willingness to apply this during class ‘named’ paragraph writing exercises. 8 My initial reluctance based on a ‘guesstimate’ of the time it would take to design, conduct, analyse and evaluate proved unfounded: the critical problem solving and writing skills of the Action Researchers made the process really easy. The concern about gaining my learners consent also proved unfounded: they demonstrated enthusiasm in both writing and assessing and have been keen to read the completed findings. My feelings are that this piece of collaborative research has improved my relationship with my students especially in justifying perceived non-syllabus teaching of study skills and examination. Analysis of the findings of this pilot study indicate that there are other factors which may have impacted on the research such as impression management, confusion about activity, timing, repetition and newness. Nor, for practical reasons are the findings representative of different academic levels, learners and curricula. But, I do believe that both the process and the product has given me a greater understanding of my own practice and my students' academic literacy needs; to the extent that I will plan to explicitly embed peer assessment throughout my teaching programme. Finally, a message to all of you who have thought about but not actioned a piece of action research – do it! It will improve your morale by providing you and your learners with evidenced based theory to justify your teaching and their learning. Bibliography Sluijsmans, D. (2002) Establishing learning effects with integrated Peer Assessment Tasks, Higher Education Academy. Sebba J, Deakin Crick R, Yu G, Lawson H, Harlen W (2008) Systematic review of research evidence of the impact on students in secondary schools of self and peer assessment. Report. In: Research Evidence in Education Library. London: EPPI-Centre, Social Science Research Unit, Institute of Education. End. Gillian Boocock, Sociology teacher Shayma Izzidien, Researcher-in-residence 5/03/09 9 Appendix 1 ACTION RESEARCH TASK Identify and briefly explain two problems of using official statistics in the study of suicide. (8 marks) – June 2007 You have 8 minutes to write your answer to this question 10 Appendix 2 Student questionnaire 1 (Pre peer assessment) Q1. How did you find the writing task? Please tick your answer on the 1-4 scale below: Not at all 1 A lot 2 3 4 I knew what I needed to write I knew how to structure my paragraph I knew how much to write I knew how to write to attract marks I knew how to write concisely 11 Appendix 3 PROPOSED COMPARATIVE MARKING TEMPLATE Name Identity number Date CRITERIA Mark awarded Words per minute (15 for AS & 20 for A2) Correct identification Correct placement of identification Correct, relevant & succinct explanation which explicitly answers the question Placement of explanation to follow identification Outline of a ‘named’ example – succinctly stated, relevant to question Placement of example to follow explanation PRE-PEER ASSESSMENT 0 1 2 POST-PEER ASSESSMENT 0 1 2 12 Appendix 4 PEER ASSESSMENT MARK SCHEME Identify and briefly explain two problems of using official statistics in the study of suicide. (8 marks) – June 2007 1. 2. 3. 4. Script clearly separates out identification and explanation identification clearly answers the question explanation clearly follows the identification and is explicitly related back to the question an example is used to illustrate the example Two marks for each of two appropriate problems identified, such as: Two further marks for each of these satisfactorily explained, such as: they may lack detail about the deceased that the sociologist considers important; they may lack detail about the deceased that the sociologist considers important: eg Durkheim found that some French suicide statistics did not record the religion of the deceased, which was of importance to his hypothesis about the causes of suicide; this makes official suicide statistics useless not all instances of suicide may have been recorded; not all instances of suicide may have been recorded: eg an attempted suicide may not come to the attention of officials because the individual survives without official intervention. Other suicides may never be discovered because the body is not found; this makes official suicide statistics unrepresentative and invalid and possibly unreliable. incorrect categorisation of cases; incorrect categorisation of cases: eg a coroner may categorise a suicide as misadventure to spare the feelings of relatives; this makes official statistics unreliable suicide statistics are a social construct; suicide statistics are a social construct: they do not represent objective reality or the real rate of suicide, but merely the number of deaths labelled as such by social actors. This makes official statistics invalid. the official definition of suicide may differ from that of the sociologist; some states may not have collected suicide statistics. Note: No marks for an identification that simply says suicide statistics are inaccurate, invalid, unreliable, unrepresentative, etc. without making clear what these terms mean, either by a definition (eg .suicide statistics are invalid, ie do not give a true picture.), or a clarification of the term through the context of the explanation. 13 Appendix 5 Student questionnaire 2 (Post peer assessment) Q1. To what extent has today’s experience of peer assessment helped you improve your writing skills? Please tick your answer on the 1 - 4 scale: No improvement 1 2 3 Much improvement 4 Peer assessment has helped me understand what I need to write Peer assessment has helped me understand how to structure my paragraph Peer assessment has helped me understand how much to write Peer assessment has helped me understand how to write to attract marks Peer assessment has helped me understand how to write concisely Q2. What are the advantages/benefits of peer assessment? ………………………………………………………………………………………………………………………………………………………… ………………………………………………………………………………………………………………………………………………………… Q3. What are the disadvantages/challenges of peer assessment? ………………………………………………………………………………………………………………………………………………………… ………………………………………………………………………………………………………………………………………………………… 14 Appendix 6 23rd February 2009 Dear students Re: Peer Assessment Research As part of my MTeach course of study I am conducting some research on Assessment for Learning. The process of my research involves: 1. creating a consensus definition of the key concept 2. identifying variables which can be operationalised 3. operationalising the variables by conducting a small observation study 4. co-authoring a report which summarises the findings and makes recommendations So, to get Stage 1 underway I would like you to consent to writing down a Definition of the concept of ‘peer assessment’ Many thanks for your help Gill Boocock 15 Appendix 7 Summary of feedback for students Research project To examine the impact of using peer assessment to improve students’ writing ability Sample and consent Eight AS Sociology students took part in the research. All students consented to take part in the research. Method Step 1: Students completed a writing exercise. Step 2: Students commented (using a score) on how they thought they performed. Step 3: Teacher marked writing exercise. Step 4: Each student peer assessed another student’s work and gave them some feedback. Step 5: Students repeated the writing exercise. Step 6: Students commented (using a score) on whether they felt peer assessment had helped them improve their writing exercise. Step 7: Teacher marked writing exercise. Findings Seven of the eight students improved their final score following the use of peer assessment. The maximum score for the exercise was 14 marks. Two students improved their score by nine to ten points, three students by seven points, two by five points and one by one point. One student’s final score however dropped by five points. Seven of the eight students felt peer assessment had helped them improve (on a scale of 1-4, in which 1= no improvement and 4= much improvement) by an average of 2.3 to 3.8. Five of these seven students showed a considerable improvement in their final score. The most common advantage of peer assessment mentioned by four students was that peer assessment helped them understand how to write better to attract marks. In relation to disadvantages, students were concerned that their peers were not qualified to mark their papers. Limitation This was a small pilot study. It is possible that peer assessment was not the only explanation for improvements in students’ final writing scores. The study should therefore be repeated with a control group to see if other variables are having an impact on the study. Full report The full research report can be obtained from Gillian Boocock. 16