Fiction writers call it writer’s block, a condition when you can’t seem to get started on a project or section of it. With writers of journal articles, it can be frustrating because they are often required to publish for promotions or tenure or job hunting. Many EFL teachers don’t even recognize themselves as researchers , so they are at a loss even more because they feel they don’t have data or a means to get it.
Let’s assume (a) you have the luxury of a little time on your hands before worrying about the submission process to a journal and (b) you have to work alone. What can you do to put together some data that can be written into an article? One approach would be to know in advance which publication you are hoping will take a paper. You can see how long it needs to be and what sort of material is typically accepted. Another approach is to consider whether you feel more comfortable with qualitative or quantitative information and how to analyze it. Even qualitative results should be examined and presented with as little bias or ambiguity as possible.
True experiments with students should involve a test group and control group, but the latter is not always easy (or ethical) to generate. With that in mind, what sort of test group can a teacher devise? A study can be about a whole class or a comparison of student groups. It can be a case study of a few subjects. A project can last part of all of a school term, or it can be a longitudinal investigation that may take longer. Within a single term, the teacher might introduce a unique lesson plan or a whole syllabus based on a pedagogical point of teaching or assessment.
Data can be extracted from interviews, survey questionnaires, samples of homework or exams, or progress reports. Each of these has its own merits and demerits. Casanave  has an easy-to-read chapter that explains how to look at writing from a reader’s point of view, what types of writing stances there are, and how to approach different types of writing.
Interviews take time, and teachers should be aware of using L1 or L2 as needed, conducting them synchronously (in person or in video/phone conversations) or asynchronously (via a written exchange), question design, and the recording and transcribing of spoken communication followed by text mining and coding of responses.
Surveys seem straightforward, but the text of questions needs to be exact to avoid misunderstanding, and ideally, they should be piloted to filter out any problems. Question design here is also important, and the teacher must ask whether an essay question could be interpreted in more than one way. Likert ranking can involve odd or even categories and may even have a neutral item such as “I don’t know” or “Doesn’t apply”. Providing sufficient response time is important, and the number of items should be gauged carefully to avoid respondent fatigue or boredom. One aspect of surveys that goes unheeded is the opportunity to circulate the responses back to the respondents and ask either for their appraisal of its accuracy or for their opinion on future steps taken from the data.
Using homework or exams may seem like the most expedient source of data since it is always available. But if previous materials are used, the teacher must remember the background leading to it, so a proper description of the “experimental conditions” can be written. If instead, a teacher plans ahead, those same conditions must still be properly designed to produce a suitable situation for the desired data. It may be possible to use materials that are already in a teacher’s collection, but care is needed to know just how it had been explained or delivered to students, and whether it is too old to be relevant.
Progress reports might be explained as student portfolios, before & after assessments  of something, lengthy data collection such as in extensive reading (ER), or evaluating the effects of course design based on a needs analysis. All of these require measured planning, since they take place over a long period of time, and any mid-course changes might introduce a disruption that causes unusable data. Such changes can, nevertheless, be explained in the discussion section of an article, but detailed records are needed to ensure accuracy.
Working with co-researchers poses its own challenges, even with the best of friends. Personalities can obstruct data collection or planning. It may be best to mutually decide in advance on roles in the work (e.g., statistician, database creator, translator, literature collector). Sometimes, colleagues have minimally active roles but provide valuable data. For example, if you compare two courses on ER in which the assessment records are collected automatically, but you depend on the two colleagues to explain ER to their students and let them read on their own. Working with partners can not only make up for deficiencies in a teacher’s own set of abilities and knowledge, but it may provide a data set from students you couldn’t otherwise access. And, it’s always nice to share feelings with someone who is well-informed about the project, instead of working in isolation where motivation and creativity might suffer.
People who are teachers only and who never take the role of the researcher may find initial attempts to collect data to be less than satisfying. That is only natural. But many who persevere have reported enjoyment in a whole new world that they never previously thought possible in their careers. Books on becoming a teacher-researcher  are growing in number.
Photo Credit: Unsplash
 Taylor, L. A. (2017). How teachers become teacher-researchers: Narrative as a tool for teacher identity construction. Teaching and Teacher Education, 61, 16–25.
 Casanave, C. P. (2009). Writing up your research. In J. Heigham and R. A. Croker (Eds.), Qualitative research in applied linguistics (pp. 288-305). Palgrave Macmillan.
 What are the types of assessment? https://www.onlineassessmenttool.com/knowledge-center/assessment-knowledge-center/what-are-the-types-of-assessment/item10637
 Shagoury, R., & Power, B. M. (2012). Living the questions: A guide for teacher-researchers. Stenhouse Publishers.
Glen served as the chief editor of the JALT CUE SIG journal for a decade. He was also a copyeditor for a science journal for five years. He has been doing scientific proofreading and editing in Japan for over 20 years, including co-editing a book on STEM English to be published in late 2021.