y attitude changed when I joined several colleagues at an action research conference in November 1997. As a tool to help teachers ask questions about their everyday work, action research promised something a little different: a chance to study my own practices and the proficiencies of my students with an eye toward what worked and what didn't. My goals were to assess the current level of performance in my classroom, experiment with new ways of doing things, measure the results, and begin again as necessary.

teach ESL at the Roger Williams Middle School in Providence, Rhode Island. This state now requires most fourth-, eighth- and tenth-graders to take part in a standards-based assessment tool created by the National Center for Education and the Economy (NCEE). The test is administered entirely in English and norm-referenced on monolingual English language users. Because of this, and because the state has mandated a 3-5% increase in each school's level of performance, I am concerned about what the consequences of this new assessment will be on non-native speakers of English. As a teacher of these students, I decided that what matters most to me can be summed up in the simple question that now forms the basis of my classsroom inquiry: Does the explicit teaching of the NCEE standards enhance ESL student performance?

ne of the basic principles of action research is that researchers need each other's ideas for stimulation and depend on other people's perspectives to enrich their own. For this reason, I elected to become part of an action research team that would apply for and receive technical assistance from an outside consultant through the LAB's Institute for Cultural and Linguistic Diversity. The group was initially comprised of all of the teachers from our district's bilingual department who had participated in the conference; but it wasn't long before our 12-person team dwindled down to just two, myself and an elementary school ESL teacher. Many of the members had joined more out of a sense of obligation to our department head than out of a desire to participate at that particular time, while others faced personal obstacles that interfered with their ability to take part. Only the two final members were involved with writing the proposal for assistance. In hindsight we saw that these factors crippled our efforts to build a larger team that could reap the greatest benefits of research collaboration.

till, our two-woman group continued to meet once per quarter to engage in dialogue about our individual questions. The contact I had with my colleague was a one hundred percent increase from the previous year and allowed me to share triumphs and concerns in a productive environment. Knowing that I would be presenting my findings to someone else also helped me to organize my thoughts and my data. Though my usual way of teaching was indeed student-centered, I came to see that it wasn't building in a circular way as I had thought it was. The increased dialogue between us contributed to the development of our knowledge about teaching and learning.

ver time, I came to see that action research demands the skills of two types of professionals: teachers who work in the trenches every day, and educational researchers who can help us to assess our teaching in a way that gives us meaningful information. Teaching is, after all, quite subjective. Our consultant helped us to become aware of the need to conduct consistent data collection from the initial stages. He also helped me to think more about the instruments of assessment I choose so that I am clearly witnessing the results of student change and not of differing conditions.

s a result, I became more consistent in the creation of tasks and the assessment of student work. For example, in a weekly computer lab each student read from Sandra Cisneros' book The House On Mango Street for a fixed period of time, summarized some aspect of what he or she had read, and related it to his or her personal life. The task addressed two NCEE standards, reading and writing. I documented student progress quantitatively and qualitatively on each element of these tasks. That is, I counted and recorded the number of pages read during the 10-minute period and the number of words written during the remaining 40 minutes. Qualitatively speaking, I evaluated students' abilities to summarize, relate the reading to their personal lives, and express their ideas in writing. I also began to record student errors in grammar, punctuation, and spelling and to use student work as the basis for explicit instruction in common areas of weakness.