THE IMPACT OF ASSESSMENT TYPE ON ADOLESCENT LEARNERS’ ABILITY TO EVALUATE THEIR OWN LEARNING PROCESS

Learning to learn, self-assessment and selfreflective skills have become a key element of foreign language learning and teaching as a result of language policy advocated by the Common European Framework of Reference (2001). The new core curriculum in Poland emphasises that learners should not only develop language proficiency and enhance communicative competence, but also acquire skills that would help them engage in self-directed learning. The ability to evaluate one’s own learning process can be enhanced by purpose-designed classroom instruction and implementation of appropriate pedagogical tools, such as formative and performance assessment. Assessment is an important element of classroom procedure as it both serves as feedback on learners’ performance and provides insights into the effectiveness of the teaching practice. Traditional tests tend to focus on highly theoretical knowledge, whereas unconventional assessment me thods attend to communicative lan guage competence, comprising not only grammatical but also strategic, discoursal and sociolinguistic elements. Moreover, these methods Anna Czura University of Wrocław Institute of English Studies ul. Kuźnicza 22, 50-138 Wrocław, Poland Tel. +48 713 752 907 E-mail: aczura@gmail.com Research interests : evaluation and language assessment, CLIL, learner and teacher autonomy, motivation, European language policy


Introduction
Learning to learn, self-assessment and selfreflective skills have become a key element of foreign language learning and teaching as a result of language policy advocated by the Common European Framework of Reference (2001).The new core curriculum in Poland emphasises that learners should not only develop language proficiency and enhance communicative competence, but also acquire skills that would help them engage in self-directed learning.The ability to evaluate one's own learning process can be enhanced by purpose-designed classroom instruction and implementation of appropriate pedagogical tools, such as formative and performance assessment.Assessment is an important element of classroom procedure as it both serves as feedback on learners' performance and provides insights into the effectiveness of the teaching practice.Traditional tests tend to focus on highly theoretical knowledge, whereas unconventional assessment me thods attend to communicative lan guage competence, comprising not only grammatical but also strategic, discoursal and sociolinguistic elements.Moreover, these methods are claimed to have a positive impact on learner autonomy (Birenbaum andDochy 1996, Fox 2008).
The subjects of this study, i.e. students of lower secondary school (aged 13-16), undergo rapid emotional and physiological changes: being aware of their own needs and capable of independent thinking, they are ready to become conscious participants of their education.Although developing autonomy and self-assessment skills at this age is difficult and requires the teacher to adjust their teaching to learners' cognitive needs, it can serve as a motivational factor and facilitate learners' personal and social development.It seems important to investigate language classroom at this level of education as there is a shortage of research concerning this age group (Jarząbek 2008).
The main aim of the one-year-long research was to determine whether the administration of alternative assessment methods (portfolio, project, peer-assessment and observation) affects learner autonomy in reference to the ability to evaluate the learning process in the classroom and in self-study situations.Should the positive relationship be established, it would enable the researcher to indicate the most beneficial assessment type for this age group in the selected aspect of learner autonomy.The study was hoped to provide a valuable descriptive insight into the language classroom and demonstrate how new assessment methods were implemented in practice.The researcher was an objective observant and did not participate in the instruction process, which helped to adopt an unbiased stance.

Research methodology
As the use of different research methods ensures validity of research results and helps to overcome the weaknesses and biases which may arise when only one method is applied (Olsen 2004), a methodological triangulation was adopted.Quantitative data concerning learners' ability to evaluate their learning process was obtained by means of a pre-and post-questionnaire prepared on the basis of descriptions of autonomous behaviour (Boud 1988, Legutke and Thomas 1991, Dickinson 1992, Breen and Mann 1997, Sheerin 1997), and acknowledged autonomy questionnaires (Michońska-Stadnik 2000, Pawlak 2004).It was administered at the beginning and at the end of the experiment to verify whether any changes occurred as a result of the treatment.The answers were graded according to a Likert-type scale where 1 indicated 'strongly disagree' and 5 -'strongly agree.' Additionally, the researcher observed the groups once a month in a variety of classroom situations: language instruction, practice, production, as well as assessment-related procedures.Finally, towards the end of the research, randomly selected subjects from each group were interviewed.The researcher's observation sheet and interview questions corresponded closely with the items in the questionnaire.

Research groups
The total of 116 learners divided into 4 experimental (groups 1-4) and 1 control group (group 5) took part in the experiment.The subjects were assessed according to uniform regulations imposed by the school, and each research group was additionally subjected to one alternative method: portfolio (group 1), project (2), peer-assessment (3), and observation (4).To ensure comparability, all groups used the same coursebook, followed the same syllabus, and were exposed to the same amount of language instruction, i.e. 3 hours per week.Each group consisted of 1-and 2-grade learners (aged 13-14).The 3-graders were eliminated to minimize the impact of maturational differences and possible washback effect of the external examination at the end of school.

Results of the questionnaire
Firstly, dependent t-tests were calculated for each group to observe whether the differences between the results in the pre-and post-questionnaire measuring their ability to self-assess their own learning process bear any statistical importance and can be attributed to the treatment the subjects were exposed to.A dependent t-test was computed by means of the SPSS 14. Figure 1 offers a visual representation of the mean scores obtained by the subjects in the preand post-questionnaire.The figure indicates that a rise in values before and after the experiment was noted in groups 1, 3, 4 and a fall in groups 2 and 5. Nevertheless, the dependent t-tests calculated for each group revealed that these differences were not statistically significant.Therefore, in the light of the quantitative data it can be concluded that the applied treatment did not affect learners' ability to evaluate their own learning process.

Results of classroom observation
To start with, the researcher focused on subjects' ability to set assessment criteria and self-assess their own progress in L2 learning.It appeared that these skills could be observed only in groups 1, 2 and 4 as the implementation of portfolios, projects and post-observation conferences were accompanied with different forms of selfassessment.Learners' engagement and ability to attend to their own performance varied significantly.Although learners in group 1 were able to establish criteria to be used to assess their portfolios, later they were applied rather infrequently and not by all learners.Moreover, despite being offered both openand closed-ended self-assessment tools, the subjects failed to attach them to portfolios.It might reflect the importance of grading as the major factor motivating the learners -as these self-assessment tools were not graded, the learners did not submit them.On the other hand, it might have resulted from the fact that learners were not accustomed to self-assessment, which is rarely promoted in traditional schools.The initial explanation of the idea of self-reflection might not have sufficed to make the subjects realise the aim and the principles underlying this procedure.
Group 2 was also encouraged to set assessment criteria of their project work, and although initially the learners were unable to use them, later most subjects gained in experience and attempted to justify their reflections on their own performance.The reason might be that the self-assessment process referred to specific tasks and was conducted under careful scrutiny on the part of the teacher, who provided direct assistance and immediate feedback.
Although no instances of self-assessment were observed in group 3, it is worth mentioning that the learners were actively involved in establishing assessment criteria to be used in peer-assessment.Although at first the learners required guidance in justifying their decisions, in later stages the majority were able to apply the standards to assess their peers' work.
In group 4, self-assessment was used during conferences concluding the outcomes of a week's observation conducted by the teacher on the basis of an observation sheet whose criteria were agreed on with the learners at the beginning of the experiment.Initially, learners were not accustomed to assessing their own performance or setting their learning goals, and only a small development was detected at the end of the treatment.It also became evident that the improvement goals the learners set with the teacher's help were soon abandoned.
The observations delineated above reveal that self-assessment in groups 2 and 4 was introduced and actively assisted by the teacher; still, learners' ability to reflect on their own performance differed.It could be explained by the fact that group 4 achieved slightly lower scores in the pre-and postquestionnaire.Furthermore, the subjects in this group had to reflect on a variety of aspects of their performance from a period of one week which might have been more abstract than attending to a tangible outcome of project work.
Next, the researcher observed learners' attitudes to their learning process and different ways of attending to their progress.Even though the teachers usually provided feedback on test results, the learners were more focused on their grades than their performance.Interestingly, in group 3, the learners were more inclined to attend to the mistakes underlined by other learners in the process of peer-assessment than on the teacher's feedback.It complies with the assertion made by Black et al. (2003) that learners are likely to pay more attention to the comments suggested by their friends rather than the teacher.

Results of interviews
The interviews suggested that the subjects were not interested in the progress they made -only a few consulted their performance with the teacher during breaks to enquire about their grades rather than learning strategies or specific language problems.Even though each teacher appointed a weekly office hour, it was treated only as time devoted to retaking tests.Post-observation conferences in group 3, which focused both on learners' ongoing performance and overall progress in L2 learning, appeared to be the only form of structured feedback involving a direct teacher-learners contact.Some learners in each group admitted that they sometimes took notes of mistakes they had made in tests and tried to correct them at home.Upon receiving a bad mark, learners wanted to retake the test; however, these plans were often abandoned.
The next questions referred to the subjects' attitude to the assessment criteria.Some interviewees in group 2 reported that although at the beginning the assessment criteria established to evaluate project work were difficult to understand, towards the end of the school year they proved a valuable help in improving the quality of work before submitting it to the teacher.The remaining subjects did not consider the criteria useful in individual work.
When asked about strong and weak points in L2 learning, almost all interviewees pointed to specific grammar and vocabulary items, which suggests that accuracy is viewed as a dominant element of L2 learning.Despite being able to indicate exact problems, the subjects could not set their own learning objectives and take appropriate corrective actions.Most of the subjects did not exhibit any long-term improvement plans but rather resorted to remedial actions implemented on an ongoing basis.Some solutions, especially in groups 4 and 5, suggest that at times the techniques or the materials the learners applied did not correspond with language problems they experienced.
A few interviewees occasionally read tasks that had been corrected by the teacher.While some learners paid attention to the faulty production, others admitted that they did it without any specific aim.These findings, however, seem to be at odds with the high results achieved in the pre-and post-questionnaire, in which the learners declared that they tried to find and correct their own language mistakes.This discrepancy shows that the declared level of autonomy does not always reflect the reality and therefore needs to be confronted with qualitative data.

Conclusions
The results of the quantitative part make it difficult to single out one method which would be most conducive to developing learners' ability to self-assess their own learning.Still, qualitative instruments revealed an emergence of diverse autonomous behaviours as a result of implementing new assessment tools.
It occurs that most subjects are able to set assessment criteria and some used them successfully to evaluate their own and their peers' tasks.The emergence of such autonomous behaviour is, to a large extent, dependent on teacher's skilful supervision and assistance.Additionally, the effectiveness of an assessment method in developing learners' ability to evaluate their own work depends on multiple factors: the nature of the task, learners' involvement as well as their attitude to the method.Some elements of the applied assessment methods failed, but it must be remembered that it was probably the first time the learners had been engaged in the processes of setting their own assessment criteria and self-assessment.It appears that to benefit from alternative assessment methods, one needs to be already equipped with a repertoire of autonomous behaviours.Therefore, the process of introducing more formative assessment has to be preceded with far-reaching changes in L2 learning: teachers need to consent to invite learners to participate in the classroom decisionmaking to enable them to make appropriate choices when working on their own.Finally, unless learner training is not encouraged in other areas of classroom practice or is not transferred to other tasks carried out at school or at home, one can hardly expect the newly developed autonomous behaviours to consolidate and become a regular occurrence affecting learners ability to self-assess their learning.

University of Wrocław, Poland
Research interests: evaluation and language assessment, CLIL, learner and teacher autonomy, motivation, European language policy

Summary
Nowadays, in order to promote language learning as a life-long process, a growing emphasis in language classroom is placed on non-linguistic aspects, such as motivation or learner autonomy.

Anna Czura
Vroclavo universitetas, Lenkija This shift of focus in language pedagogy has been reflected in documents regulating language policy in Europe and in Poland.Learner autonomy can be fostered in the classroom by means of various pedagogical procedures, for instance alternative assessment methods, such as portfolio, project, observation, and peer-assessment.The research presented in the article aims to investigate whether and to what extent the abovementioned assessment methods affect one aspect of learner autonomy, namely learners' ability to evaluate their own learning process.This ability is particularly useful in developing learners' ability to self-reflect on their own learning, analyse their needs, set learning goals, and implement necessary remedial actions.The research was conducted in a lowersecondary school over the period of one school year and is based on methodological triangulation comprising both quantitative and qualitative data elicitation tools: a questionnaire, classroom observation and interviews with the learners.The application of diverse instruments helped to obtain an in-depth overview of the impact language assessment exerts on the dependent variable.Although the statistical analysis failed to prove that the administration of alternative assessment in the classroom affects adolescent learners' ability to evaluate their learning process, the remaining research instruments indicated some traces of autonomous behaviour that evolved in the subjects in the course of the experiment.The research also allowed the author to observe and analyse different problems the teacher faced in the process of introducing new assessment methods and to formulate a number of practical implications concerning fostering autonomy in a language classroom.KEY WORDS: learner autonomy, alternative assessment types, self-assessment, adolescent learners.

Figure 1 .
Figure 1.The mean results obtained in the pre-and post-questionnaire measuring self-assessment skills