TYPICAL METHODOLOGICAL FAILURES DEMONSTRATED BY STUDENTS OF TEACHERS ’ COLLEGE IN CONDUCTING CLASSROOM ACTION RESEARCH

Conducting a research is a compulsory skill to be accomplished by any teacher students of English language teaching (ELT) department in teacher’s college. As a minimum requirement, each teacher-student must have experience to conduct a sophisticated Classroom Action Research (CAR). Within this skill, a teacher-student is expected to grasp various challenges that may occur in learning, therefore to take necessary actions in order to improve the quality of learning. Hence, this paper is composed due to an assumption occurred in a preliminary study that students in English language teaching department have demonstrated numbers of methodologically failures in conducting CARs. This assumption stimulates a question to answer further, “What sorts of methodologically failures are demonstrated by teacher’s college students in conducting CARs?” Researchers as the main instrument in this study develop theoretical criteria form based on prominent works of Susanto (2010), Hult and Lennung (1980), McKernan (1991), Kemmis and McTaggart (1992), Winter’s (1996), andMcNiff (2002) to frame the analysis. Data is taken from selected works stored in a library of a teacher’s college in Nusa Tenggara Barat, by employing two criteria to students’ works (mini-thesis), i.e. year of publication is 2015 and marked with ‘A’ from internal examination board of the college. These criteria constrain only two scripts to be further analyzed. By applying content analysis, this study reveals various kinds of methodological failure in students’ works about CAR, i.e. (1) violation to collaborative principle of CAR, (2) violation to the four characteristics of CAR (situational, participatory, evaluative, and cooperative), (3) the use of learners’ achievement as benchmark of success, (4) developed in quantitative study, (5) failure in positioning the researcher as a teacher or collaborator, and (6) failure in the construction of


INTRODUCTION
The need of professional teaching squad becomes one of the most vital matters in pursuing the primary goal of education.Teacher is one of the main components of formal education realm.Hence, government has issued various policies to improve the quality of teachers.However, the policies are more allocated to on-duty teachers.This paper begins with a more fundamental issue by assuming that the future teacher education should be also accommodated the grand design of national education policy.One of primary skills ought to be acquired by the future teacher is research skill, which can support his future carrier as teacher.As a teacher, someone should be able to conduct a sophisticated classroom action research (CAR), which can support him/her in their future jobs.Therefore, this paper is intended to examine the teacher students' competence in conducting CARs as part of their college activity.

Literature Review
In 1940s Kurt Lewin first introduced action research attracting social scientists with its liberating intent to go beyond various field of interests.This approach ever since has been widely used to integrate two kinds of work, i.e. research and action.Principally, scientists have another strong argument on how a research can bring such an immediate solution to certain problems.Cohen, et.al. (2007) preview an action research as a powerful tool for change and improvement at the local level (297).Holly and Whitehead (1986) mention how this tool is applicable in almost any setting by counting two basic considerations to generate it, i.e. problem (relating to people, tasks, and procedures) and change (for desirable outcome).In terms of education, Holly and Whitehead further suggest the implementation can be taken by a teacher alone, a group of teachers, and a teacher(s) with research fellow(s) as an outsider.
Reason and Bradbury ( 2001) define action research with some keywords, i.e. participatory, democratic process, human purposes, action and reflection, theory and practice, and practical solutions.These keywords cover the basic principles of conducting an action research, in which it begins with identified problems and pursuing for an effective solution of it.Earlier scholars suggest quite similar account to define an action research, such as (Hopkins, 1985;Ebbutt, 1985;Cohen and Manion, 1994;Corey, 1953;and Kemmis and McTaggart, 1992).Winter's (1996) suggests six key principles of action research, i.e. reflexive critique, dialectical critique, collaboration, risking disturbance, creating plural structures, and theory and practice internalized.All of the scholars are linked with a keyword, i.e. improve, in which change is expected to enhance the pursuit of institutional goal (or in this study classroom goal) in varied social entity, including classroom as a small unit or beyond it.
Thus, the study perceives a CAR as designed change conducted in the level of classroom to improve the process of pursuit the learning goals which identified as problem triggering the action.Susanto (2010:17)  Apparatus requires teachers to have scientific publications to pursue for a higher position in their carrier.Teachers are expected to publish at least some CAR reports within their professional experiences.In this sense, every teacher is encouraged to develop research skill.Minimum requirement of education background (in this case a degree in education) also implies that every future teacher should be well equipped with this skill.Therefore, teacher's college should be able to develop their teacher students' skill in research.Based on this fact, this study is designed to be an evaluative work to give further input for teacher's college management board in order to improve the quality of their outcomes.
A preliminary study towards some students' works labeled as CAR leads to an assumption that teacher students in college demonstrate methodological failures in conducting CARs.This assumption is also aligned with the fact that many teachers dispute their incapability to conduct sophisticated CARs.Therefore, this study is intended to conduct a diagnostic work to determine the parameter of failures demonstrated by teacher students.The only research question in this study is, "What sorts of methodologically failures are demonstrated by teacher's college students in conducting CARs?"

METHODOLOGY
This study employs content analysis (CA) using a frame that consider critical notation of CAR by Susanto (2010), Hult and Lennung (1980), McKernan (1991), Kemmis and McTaggart (1992), and Winter's (1996).Cohen et.al. (2007:475) provide a simple definition of CA as an effort to summarize and report written data covering main contents of data and their message.This analytic tool enables scholars to break in the entire parts of manuscripts.Two selected works of teacher students are taken as sources of data.The selection of data sources are following two researcher-driven criteria, i.e. year of publication should be 2015 and the works has got 'A' mark from the internal ad-hoc examination board.Based on internal data in the teacher's college, two manuscripts are then used in further analysis.Both works are then investigated through CA by employing theoretical frame as stated earlier.In analyzing data, this study follows the three-step approach postulated by Miles and Huberman (1994), i.e. data reduction, data display, and drawing conclusion.About the use of CA, it helps the researchers in making inferences by systematically and objectively identifying specified characteristics within a text (Stone, 1966:5).Through this analysis, one can explore his or her critical analysis on specific content of discourse, in this case a research report.The analysis must be performed relative to and justified in terms of the context of the data (Krippendorf, 1980:23).

Findings
In data analysis, this study applied an instrument to analyze two manuscripts that were selected based on categories mentioned in earlier section.All data collected in this process were displayed in table 1 and 2

DISCUSSION
Susanto (2010) constrains the steps of developing a CAR as three-cycled activity, i.e. preparation, implementation and observation, and reflection.Hence, by conducting CA to two manuscripts, this study identifies three categories of failures, pre-action, action, and post-action.The categories were developed based on sequences of CAR.Each category contains several types of failure within the sources of data.

Pre-Action
Pre-action phase refers to a stage of a CAR covering the following issues:

Finding research collaborators
Susanto (2010) distinguishes two types of participants in a CAR, i.e. initiator and partner researchers.The former refers to an expertas an outsider, while the lateris a teacher as an insider withadequate knowledge and competence relating to course subject and research issues.In this sense, both writers of the manuscripts have taken positions as an outsider.

Identifying, evaluating, and formulating research problem or issue to be further solved
A CAR is supposed to rise from reality that occurs in a classroom emerging during teaching learning process.The problem is constrained to element(s) of interaction,in which a CAR is encouraged by dissatisfaction towards the quality of learning outcomes and the willingness to improve it (Susanto, 2010).Susanto further suggests 7 categories recognized as the sources of problems, i.e. indicator, learning materials, learning strategy, instructional steps, assessment, teaching media, and learning sheet.To respect Susanto's, this study traces several typical failures in the two manuscripts.
It is found that problem in the two manuscripts tend to emerge based on the outsider's subjective assumption and not elaborating it from the teacher as the authoritative person in the class.Obviously, the design of the CAR is not classroom based research as suggested by many experts.This entraps the researcher to bring in something from the outside.In the beginning, manuscript A mentions a boring classroom, which is quite relevant to urge a CAR.But later, the writer takes a leap by forwarding students' vocabulary mastery as the core problem of the study.In this sense, the writer has switched from issues of process quality to learning achievement.The later tends to be recognized as a quantitative inquiry that is more relevant to an experimental study.Manuscript B has demonstrated similar tendency.First, the writer has noted several issues related to students' classroom performances, i.e. feeling shy, fear of making mistake, difficult in expressing their idea, seldom to practice, minimum repertoire of vocabulary, misunderstand the given material, and lack of confidence.He further formulates an assumption regarding the students' performance, which can be seen as working hypotheses, that covers students' motivation, attractiveness of learning material, and technique in teaching.Moreover, he underlines that teacher tends to count heavily on writing than speaking.This seems to be irrelevant with his earlier notes.Later, the given research problem becomes irrespectively to the earlier arisen problem in the introduction of his writing.
Statement of problems seems to be detached from the classroom context bybringing inout-of-context issues based on the writer's justification.In manuscript A, the writer underlines the classroom situation by defining the students' low motivation in learning activity, which he labels as 'bored'.Unfortunately, the term 'bored' is not emerged as an element in the formulation of research problem.The formulation seems to be more focus on examining a newly teaching model, i.e. pocket chart, to improve students' vocabulary.Based on this formulation, the CAR fails to place the classroom based issue as a departing point in developing the research design.After reviewing manuscript B, this study finds that the writer fails to bring in all problems arisen in the observed class into a proper research problem.All problems are simplified as students' speaking skill.Thus, he offers a different kind of media as a treatment to the given problem, which is marked as not situational one.Thus, these two notions become the variables mentioned in the research problem.

Formulating working hypothesis
While both manuscripts indicate misleading in generating research problems, both of them also fail to generate appropriate working hypotheses to be further conducted within the studies.The failure in drawing proper working hypothesis in a CAR may lead the writer to wrong direction.
Manuscript A mentions about students' low motivation in learning, but none of working hypothesis discusses this fact as the main reference for the writer to formulate further action in future lesson plan.Manuscript B also asserts some learning problems dealing with students, but the writer fails to produce relevant working hypothesis.He simplifies the problem as speaking skill issues,and then he proposes a new media assumed as more effective in improving students' speaking skill.

Numbers of participants
Both writers in their manuscripts engage only two participants in their research designs, one as a teacher and another as an observer.Susanto (2010) suggests three participants as the minimum numbers of participants including the teacher and two observers.By employing more than one observer, it may reduce the potential researcher subjectivity in giving justification.A failure to present minimum subjective claim within a CAR can decline the quality of the work.In this regard, both manuscripts impose two likely subjective claims.This circumstance is unfavorable in a scientific work, in which the only observer makes his judgment within his mind and modified by individual bias.

Role of participants
Manuscript A indicates how the writer takes position as a teacher rather than an observer.Meanwhile, the teacher of the class is assigned as an observer.This positioning has several impacts, i.e. ( 1) by design it can interfere the nature of the classroom by placing an outsider to lead the teaching learning process.Susanto (2010:25) asserts that the real teacher should play his natural role to lead classroom interaction by using modified lesson plan as his guidance.He further argues that the teacher has known well the setting, coped and comprehended the learning materials, as well as his best in identifying and recognizing the pupils.

Relevancy of literature
Different to other research report, in terms of the state of the arts or theoretical frame of a CAR is not necessarily too comprehensive, as suggested by Susanto (2010).He argues that teacher limited access to the recent theory of foreign language teaching loosen the teacher researcher from the obligation to compile a sophisticated review of literature.However, in the case of involving an expert as the outsider in the research, a CAR is better to be equipped with the complete ones.

Ongoing theoretical response to the given problem
Failures in formulating the proper research problems and working hypotheses in their works also cause improper temporal response in their chapter two, i.e. review of related literature.In manuscript A, the writer proposes chart pocket and vocabulary mastery as the two core theoretical responses to the given problems in chapter one.There is no literature discussing about learning motivation or 'bored' students.Meanwhile, manuscript B which defines various learners' problems simplified as speaking skill and offering English movies as the solution, discusses too broad theoretical framing in chapter two, covering language as communication device, language elements and skills, speaking as language skill, and learning media to facilitate foreign language acquisition.No review is provided regarding learners' problems as stated in the introduction.These facts are driven by the writer's misidentification of sources of problems.In other words, regarding the aim of chapter two, both writers fail to give adequate state of the arts of their works.

Modified lesson plan and problem solving submission
Susanto (2010) emphasizes that a CAR should be departed from the existing lesson plan, which was modified or revised in order to improve the quality of learning.Both manuscripts indicate opposite direction to this principle.Hult and Lennung (1980) and McKernan (1991:32-3) assert that a CAR should take feedback from earlier cycle as consideration in improving learning, which is also suggested by Kemmis and McTaggart (1992).
Manuscript A fails to demonstrate the process of revising the original lesson plan used by the teacher in his instruction before conducting the CAR.The writer seems to bring a new lesson plan detached to the existing one, which marked as cycle 1.The same thing also emerges when he moves to cycle 2 after justifying that cycle 1 has failed to solve the problem.The same indication also appears in manuscript B. First, the writer does not depart from the original lesson plan used by the teacher before the research.Hence, the writer comes with a totally different lesson plan offering brand new media of teaching.He conducts the research in two cycles too, but the second cycle is only repeating the same actions in cycle 1, with no revision at all.Kemmis and McTaggart (1992) propose a small-to-large protocol of change due to classroom situation as constrained in the existing lesson plan.This may avoid any researcher to bring in the out-of-context ideas into classroom.Both manuscripts fail to obey this principle by adopting brand new teaching media.

Standard of success
Many experts have suggested the boundary of a CAR defined within the process of learning in terms of quality by departing from the existing lesson plan.The success of a CAR should also begin from this notation rather than just seeing students' learning achievement simply through an achievement test.Both manuscripts explicitly indicate the use of student's scores towards a prepared test as the primary standard of success of actions in the CARs.Susanto (2010) suggests this kind of information as additional information of a CAR, not the primary one.Therefore, the test results cannot be used as hints for the writers to further decide whether they need to proceed to the next cycle or terminate the research.Based on Hult and Lennung (1980) and McKernan (1991:32-3), a CAR is aimed at advancing the quality of human actions.The quality cannot be defined by testing participants' (learners) comprehension.They further suggest that this kind of research should only concern about abrupt issues in classroom.
While, manuscript A mentions the use of subject passing grade (Indonesia: KKM), manuscript B simply uses an achievement test results with no explicit passing grade to determine the success of learning.The writer of manuscript A conducts pretest-posttest approach in his effort to give judgment whether his action effective or not.Yet, he does not conduct a further statistical analysis of both tests.Manuscript B only indicates that the students' average scores are showing improvement from 44.55 to 75.38 as the only reason to terminate the action.Kemmis and McTaggart (1992) mention the selfreflective spiral as one of the key principle to conduct a CAR.In other words, the decision making is based on the result in reflection stage, not through a test.Hult and Lennung (1980) and McKernan (1991:32-3) also state that a CAR should be formative, in which a researcher makes a formative judgment regarding the quality of actions.Thus, it becomes contradictive if a CAR uses achievement test as a measure to determine whether actions have met the requirement.Both manuscripts fail to demonstrate critical analyses as suggested by Kemmis and McTaggart (1992), they only consider students' test results as the standard of success.Winter (1996:13-14) mentions that a CAR should create plural structures rather than a single authoritative interpretation.This principle allows every participant to give their accounts and critics.Unfortunately, both manuscripts in this study have violated this principle by only using test result as the primary consideration.

Developing research instruments
In terms of research instrument, most experts suggest the use of field note as a tool to collect data from classroom.However, none of the two manuscripts contain field note or sample of applying field note in the sites.
Furthermore, this study reveals that both manuscripts use tests as instruments.Between both writers, the difference is in the type of test.Manuscript A applies written test (multiple choice and open-ended test), while the other manuscript prefers using oral test focusing on five components (vocabulary, pronunciation, grammar, fluency, and comprehension), besides using an interview guideline that consists of 18 questions set in three categories (family, movies, and pets and animal).The interview guideline has no clear relationship to the issue being discussed in all chapters of the manuscript, especially to the core problems of the study.

Research Design
Experts suggest any researcher to develop a CAR as a descriptive qualitative study.Hence, a researcher should use qualitative instruments of data collection.Therefore, the data analysis technique is developed in adequate manner, i.e. four-cycle data analysis, i.e. planning, action, observation, and reflection.CAR is different with other qualitative research in terms of this fundamental design.Both writers fail to go along with the cyclical process of CAR as the ultimate inquiry tool.It is obvious that both writers only use the CAR format in their proposal, but not implementing the truly CARs in their studies.

Procedure of entering the site
One of important steps in conducting a CAR is strategy of entering the site.This step helps any researcher to reduce potential bias caused by the presence of outsider, in this case the writer.Thus, Susanto (2010) suggests any researcher to have at least one or two preliminary visits before starting data collection.The frequency can be less or more based on observer's judgment.He can begin to collect data once he convinces that the classroom interaction has run naturally and ready for data collection procedure.Both manuscripts do not give clear steps of entering the sites.So, the validity and reliability of data used in the two manuscripts are questionable considering no justification that classrooms are naturally entered by the two writers.

Preliminary activity
Susanto ( 2010) McNiff (2002:71) suggest simulation as part of conducting a CAR.It may help a researcher to overview the strength and weakness of a design, in order to make necessary improvement and to ensure all participants in coping with particular roles during the actions.In the end, it can reduce misleading implementation of actions conveyed in lesson plan.Teacher and observers should be well understood about the entire steps.In manuscript A, the writer discusses the task given to the observer.However, he misunderstands the proper positioning by assigning the teacher as an observer.Meanwhile, in manuscript B, there is no explanation about conducting a simulation session with the other participant indicated as failure procedure of conducting a CAR.

Conclusion
CAR is a kind of research aimed at increasing the quality within learning.Based on the given explanation, there are some failures found in the two manuscripts caused by misunderstanding the CAR principles, i.e. procedure of problem identification, selecting participants and distributing tasks, using pre-test and post-test, focus on learning achievement rather than learning process, bring in new things into the classroom, no clear path of lesson plan development, improper design and use of data collection instruments, and there is no well prepared pre-action activity.Various typical failures may result unqualified manuscripts, which can also misled the student teachers in applying CAR in their future carrier.However, teacher students are not solely responsible for the failures, since both of them have been supervised by some faculty members.This indicates that the failures can also be reduced through an intervention to the faculty members.

Suggestion
Following the conclusion above, this study suggest some important issues: 1) any researcher must be better in understanding the principles of conducting a CAR, 2) faculty members need to enrich their scientific perspective while supervising teacher students in conducting a CAR, 3) further study can examine certain subject related to research skill, and 4) chairman of department should consider to improve the quality of curriculum as well as individual capacity of each faculty member in order to improve the quality of student's research in the future.