5. DISEÑO DETALLADO _____________________________________ 29
5.4. Análisis de vuelco
A recurrent argument in the literature on assessment is that teacher evaluation should draw on multiple sources of information (Grissom and Youngs, 2016). Introducing collaborative teacher assessors to the collection of data creates increased opportunities for triangulation, examining reliability, and fit statistics. Another source of potential evaluation data is students themselves. An integral component of assessment for learning practices is self-assessment, which is concerned with revision and improvement. It enables students to independently assess their own and other students’ progress with confidence, rather than always relying on teacher judgment (Benson, 2013). When students self-assess, they are actively involved in the learning process, and their independence and motivation are improved. A great deal of
research has illustrated the educational benefits of self-assessment, with calls that self- assessment should be included in every major assignment a student undertakes (Wiggins, 1993).
Boud and Holmes (1981) highlight these benefits as threefold: firstly, improving feedback to the students by increasing its provision during the learning process in a more timely manner; secondly, providing additional reinforcement of teaching aims and objectives, with the criteria and materials undergoing consideration several times more than usual; and finally, self-assessment is one of the few teaching strategies that allows a reduction of teacher workload, in conjunction with a corresponding increase in educational benefits. McDonald and Boud (2003) point out that students themselves experienced the benefits of self-
assessment: the vast majority found it empowering; that it gave them greater independence; improved their analytical and critical skills and general studying habits; and allowed them to prepare more effectively for their final exams. Students who had undergone self-assessment also outperformed similar students who had not experienced such assessment procedures. An
additional benefit was that it allowed students to develop the capacity to become life-long learners (Boud and Falchikov, 2007).
Nevertheless, some teachers have raised concerns in regard to self-assessment. Boud (1999) explains that care is needed if self-assessment also becomes a form of formal
assessment. He goes on to propose that this form of assessment should focus primarily on being part of the learning process rather than as a substitute for other types of assessment. Similarly, for many teachers, the value of self-assessment is often questioned; the reliability of such assessment is often not trusted, and its validity questioned (Noonan and Duncan, 2005). Therefore, it is often absent from many syllabi and classrooms, as these issues may appear to many to negate any potential benefits, particularly when the great deal of time and effort required to produce meaningful self-assessment procedures is considered.
However, to regard self-assessment purely in terms of the learning process without maintaining a focus on actual grades, or to discard it entirely on the grounds of validity and reliability issues, ignores important potential benefits. By examining the correlations or differences between student and teacher grades, it is possible to increase understanding of issues with syllabus design or teaching techniques, and to emphasize opportunities for effective feedback on an individual level. Exploring the issues of validity and reliability in a positive manner, and exploring differences in student and teacher grades, could impart valuable knowledge about students’ perceptions of their spoken performances. The first of these benefits is the potential to highlight opportunities for targeted feedback.
Sadler (1989) and Butler and Winne (1995) have highlighted the need for strategies to provide higher-quality feedback to students, and to encourage students to develop less
reliance on teachers for their learning. McDonald and Boud (2003) have argued that formative assessment does not provide sufficient opportunities for feedback, beyond that which is the sole province of the teacher. They therefore argue that students themselves have an essential role to play in shouldering the responsibility for assessment. According to Black
and William (1998), the gap must be closed between students’ lack of knowledge, as revealed by feedback, and the teacher’s aims and objectives. Students following a prescribed dictation of assessment by the teacher without understanding its purpose are unable to learn. In this regard, Black and William (1998) describe self-assessment as an essential component of learning. Self-assessment provides an opportunity for highly detailed feedback, closing the “gap” in knowledge that may exist between teacher and student expectations, which was highlighted in Cycle One.
Of course, more accurate self-assessment between student and teacher would allow feedback to be a continuous process during the learning experience, rather than a summative process after the assessment event. If students are well aware of the criteria before assessment and these criteria are unambiguous, with classroom aims and objectives made clear, students’ self-assessment accuracy could be improved. Boud (2000), like Ross (2006), describes the necessity for a criterion-based framework, so that learners assessing themselves are aware that their achievements are the result of meeting an acceptable standard, as opposed to simply doing better than other students. Brown (2004, p. 19) also highlights the need for an
adequacy of construct definition in tests of second-language communicative competence. Brown explains that communicative competence is often “an abstraction that is rarely defined with any precision in terms of actual test performance”. Citing Bachman (1990, p. 50), he quotes that in order to “maximize the reliability of test scores and the validity of test use, we should ... provide clear and unambiguous theoretical definitions of the abilities we want to measure”.
Some research has explored the accuracy of student self-assessment. According to Blanche and Merino (1989), although errors do occur, considerable research has shown that learners can be accurate in assessing their own abilities. Blanche and Merino claimed that the key to consistent overall agreement between self-assessments and rating is that the skills to be assessed in foreign languages should be clear and detailed, echoing the views of Boud and
Falchikov (2006) and Ross (2006). In a range of different educational contexts, Black and William (1998) found correlations between self-assessments and teachers’ assessments of 0.71; and this success was mirrored by Merrett and Wheldall (1992), Griffiths and Davies (1993), Powell and Makin (1994) and Meyer and Woodruff (1997). Boud and Falchikov (1989) reviewed 48 studies reporting self–teacher assessment agreement, and found that self- assessments agreed with teachers’ ratings in most cases. They did, however, highlight numerous shortcomings regarding the quality of many of the studies, and found that there was extensive variation in what constituted student and teacher agreement. Other research has highlighted issues with the accuracy of self-assessment. Ross (2006) found mixed results for self–teacher agreement in studies of second-language learning, with wide variation between studies, and a mean correlation of .64. It has been argued that in the field of language learning, there has been a failure to examine student self-assessment in relation to criterion- referenced assessment, whereby the criteria used by teachers and students are adequately defined (Ross, 2006).
The aim of this study, therefore, is to generate clearly defined criteria based upon theoretical understandings of language performance. These criteria are to be made clear and quantifiable to students. Feez and Joyce (1998, p. 2) argue that the foundations for specifying abilities and outcomes in a genre-based approach are laid in the course design: “learners cannot learn everything about a subject at once nor can they learn effectively from a random collection of unrelated items. Therefore the teacher needs to develop a systematic plan for course content which will lead to desired outcomes”. They explain that this systematic plan is the “syllabus”, which provides a map for the teacher and the learners, and is based upon explicit objectives: “it is a public document, usually prepared by teachers and negotiated with learners. It specifies what is to be taught in any particular course of study”. Self-assessment procedures were integrated during the planning phase of Cycle Two by introducing a self- assessment component to the final assessment, as well as weekly self-assessments of student
performance during each class, to be recorded in their student journals, as described in section 4.6.3.
6.4 Ethical Considerations
As stated in section 4.4, the PhD Student Research Ethics Approval Form (REC1) was submitted twice: this included the submission for data collection in Cycle Two, taking into account additional student- and teacher-generated data.
6.5 Action: Data Collection
Data collection followed the same pattern as Cycle One, as illustrated in Table 4.2; but with additional data collection points for Cycle Two, based on the inclusion of quantitative student self-assessments, qualitative student self-assessment reflections; and data from collaborating teacher researchers via teacher journals, weekly research meetings and quantitative data collection of assessments. As opposed to the 27 students of Cycle One, Cycle Two reflects data collected from 240 students. This section will explain these additional data collections in more detail.