Figure 1 illustrates modes of research across two dimensions: qualitative/ quantitative and insider/ outsider. Action research usually falls in the lower-right quadrant of the matrix: qualitative research by insiders.

Such "insiders" may be preschool teachers, assistant principals, reading specialists, high school math teachers, curriculum coordinators, coaches, university professors -- any of us who study our own practice as educators. We find the ethical safeguards of the outsider doing quantitative, experimental research (random selection, control groups, removing the personal influence of the researcher) either irrelevant or problematic for us as insiders. In the same way, the ethical safeguards of the outsider doing qualitative research (anonymous informants, disguised settings) are subverted as soon as the inside author is named; in addition, anonymity may defeat the insider's goal of open communication with students, colleagues, and parents. Ethical guidelines need to be rethought for the special case of research by practitioners in their own workplaces.

When does good teaching become research? The line is hard to draw until a study is underway. Research tends to involve

1. more systematic documentation and data gathering
2. more self-reflection in writing
3. more audience (collaboration, presentation, publication)

It is this third feature that most often brings ethical dilemmas. If our journals remain private and our videotapes are not played, we can inquire with equanimity. But though we document our own practice, we rarely work in isolation. We draw energy and insight from colleagues, seminars, outside researchers. From this informal sharing, we may reach out to wider audiences at conferences or in print. Dilemmas of responsibility and ownership arise, and the academic codes of conduct are silent.

An Alternative Human Subjects Review?

This guide emerged from discussions in the Teacher Educator Seminar of the Action Research Collaborative in St. Louis, MO. We meet monthly, a dozen or more faculty from several colleges and the Gateway Writing Project, to share our own action research and to discuss issues in facilitating teachers' work. As some teacher inquiries led to proposals for grants and dissertations, I found myself struggling with the language and assumptions of the mandated "human subjects" review (see P. V. Anderson, 1996). "Human" in contrast to what? Rodent? And "subjects"? How can this impersonal term identify the relationship between teacher and student? I raised the issue of ethics at the seminar and found that others had similar concerns. Could we, perhaps, revise the university's IRB process to make it appropriate for action research?

An ARC committee examined human subjects reviews and ethics policies from local universities and from the American Anthropological Association, the Oral History Association, and the American Educational Research Association. I began writing the guide, bringing drafts to ARC seminars for discussion. Feedback came from a wider audience of teachers and administrators in conferences, graduate courses, and teacher research groups.

What at first seemed a rather straightforward exercise in translation proved a formidable task. The more I tried to address the different contexts and communities in which practitioners do research, the more convoluted our ethical guidelines became. As teacher educators, we began to see that a "new paradigm code of ethics" would itself become "procrustean" (Gregory, 1990, p. 166). I finally abandoned the goal of an alternative "human subjects" review. Even if an adequate ethical code could be written, having it enforced by administrators who were not grounded in action research might do more harm than good. How, then, could we protect the rights of students without inhibiting the rights of teachers to gather and reflect on data from their own classrooms?

The document contained in this chapter supports what Jennifer Mason (1996) calls an "active and self-questioning" approach (p. 167):

It is because of the complexities of research ethics, and because there is unlikely ever to be one clear ethical solution, that a practical approach to ethics which involves asking yourself difficult questions--and pushing yourself hard to answer them--is particularly appropriate. (p. 29)

The guide offers a set of questions as a heuristic for reflection. A research team or dissertation committee can work through the guide with anyone who is planning a project. Most questions ask the researcher to discuss a potential ethical problem, to consider alternative actions, and to explain his or her choices.

The following "Questions for Review and Reflection" use the categories of a typical IRB ethics review only as a point of departure. Part I requests an overview of the project. Part II examines the role and location of the researcher within the research setting. Part III asks about methods, and whether the research falls within the everyday decisions of a teacher or whether there is some further intervention. Part IV examines what the IRB calls subjects and the impact of the research on relationships with students, colleagues, and administrators. Part V considers the consequences of the research for participants and ways to reduce risks either through informed consent or through openness, dialogue, and acknowledgment. Part VI deals with publication and the issues of credit, privacy, voice, and multiple perspectives. Part VII poses some questions that have been generally ignored in IRB documents, but have been especially problematic for researchers studying their own practice.

QUESTIONS FOR REVIEW AND REFLECTION

I Overview of the Study

1. Briefly describe your project as you see it today.
2. What is your time frame? Is this a one-shot project or do you anticipate several cycles? Have you done a preliminary study?
3. What problem does your research address? Is it a problem in your own practice? Or is it a problem with your students or with your administrators? Who owns the problem?
4. What (initial) action will you take? What do you hope to accomplish?
5. List your research questions as they appear at this time. (Questions will be revised or refocused during your project.)

Comments. This section can help beginners grasp the mind-set of action research, recognizing that their study will not provide simple "yes" or "no" answers, that each cycle of research will lead them to a deeper inquiry, and that the questions themselves are likely to change. They may need to redefine their initial problem so that they "own" it and can therefore take action to address it. Hubbard and Power (1993) portray this initial phase of research:

As Glenda Bissex writes, "A teacher-researcher may start out not with a hypothesis to test, but with a wondering to pursue" (1987, p. 3). All teachers have wonderings worth pursuing. Transforming wonderings into questions is the start of teacher research. (pp. 1-2)

II Location

1. Are you, the researcher, also a participant in the setting where this research will take place? Specify your role (teacher, supervisor, principal, counselor, social worker, and so forth)
2. Map your position on such dimensions of culture as gender, race, age, region, ethnic heritage, education, class, and family.

Comments. Practitioner research is inquiry into one's own practice. Therefore we begin by looking at ourselves and what we bring to the research--personally, culturally, and professionally (see Zeni & Prophete, Chapter 10 in this volume). Traditional experimental research sees the ideal researcher as neutral, unbiased, objective, but as teachers we are personally involved in our classrooms. Instead of trying to distance ourselves, we need to articulate our own position and analyze how it differs from that of other participants in the research.

As Gesa Kirsch (1999) explains, "The goal of situating ourselves in our work and acknowledging our limited perspectives is not to overcome these limits--an impossible task--but to reveal to readers. . . what factors have shaped the research" (p. 14).

III Methods

1. For this research, will you gather data on your normal educational practice and on changes in curriculum, instruction, and assessment that you could make in your role according to your own professional judgment?
2. List the kinds of data you plan to collect (e.g. fieldnotes, taped interviews, writing samples).
3. How is this plan different from the way you normally document your practice? Consider two or three alternative ways you could gather data for this project. What are the ethical implications of choosing your preferred method?
4. At this time in your research, what do you aim to understand? What do you aim to change?

Comments. Traditional research in education is conducted by outsiders who intervene in the instructional process to answer questions that may benefit themselves or the profession in general. While the goal may be to improve instruction, rarely do the teachers or students in the study benefit directly.

In practitioner research, insiders study their own professional practice and frame their own questions with an immediate goal to assess, develop, or improve their practice. Such research belongs to the daily process of good teaching, to what has been called the "zone of accepted practice."

This concept has been used to determine whether research requires a formal proposal to an institutional review board (IRB). Answering "yes" to question 1 made a project exempt from full review by a university or district IRB. Most practitioner research would thus be exempt.

In 1995, the U. S. Office for Protection from Research Risks tightened its regulations, directing that all research, even in the zone of accepted practice, must be reviewed (P. V. Anderson, 1998). This directive poses an ethical dilemma for educational institutions that want to encourage teacher research. One solution is an expedited review of projects conducted within an established seminar or professional development partnership (see chapters 1, 5, and 12 in this volume). Another is a reviewer who understands qualitative research and is designated by the IRB to offer quick, appropriate feedback to practitioners.

Beyond the institutional regulations, research in the zone of accepted practice may involve genuine risks to participants. Practitioner research should be discussed with a principal, supervisor, or other school-based mentor and often with parents. It must, of course, conform to local school policy. A research team based in the school, district, or university, with advice from a professor or team leader, can help members work through the ethical dilemmas that may surface during the research.

Question 4 takes a closer look at purpose as well as methods. Traditionally, subjects were not considered at risk if the research was merely unobtrusive observation of behavior not "caused" by the researcher. But a teacher inevitably tries to cause some outcome (such as learning); action research involves taking some action and observing what happens.

"Understanding," however, is a valid goal for practitioner research. We may cause more harm than good by trying out new methods to solve an alleged problem we have never taken the time to understand, document, and analyze. The first cycles of action research might well aim simply to understand the problem; later cycles might involve specific changes designed to address it.

IV Subjects, Subjectivity, and Relationships

1. Describe the individuals, groups, or communities you expect will be touched by your project. List their roles (student, parent, resource teacher, and so forth). Which participants are minors?
2. Analyze the power relations in this group. Which people (e.g. students, parents) do you have some power over? Which people (e.g. principals, professors) have some power over you?
3. What shared understandings do you have with these people? Do you have personal bonds, professional commitments? Will your research strengthen this trust or perhaps abuse it?
4. Will your study attempt to read and interpret the experience of people who differ from you in race, class, gender, ethnicity, sexual orientation, or other cultural dimensions? How have you prepared yourself to share the perspective of the "other" (coursework, experiences, other sources of insight)?
5. Will an insider review your questionnaires or teaching materials for cultural bias? Have you provided for consultation by adult members of the community? How will you reduce or correct for your misreading of populations who differ from you?
6. Does your inquiry focus on people with less power than you? Children in classrooms are always vulnerable--especially if their families have little money or education. ("Where are the ethnographies of corporate boardrooms?" asks House, 1990, p. 162.) How does your project demonstrate mutual respect and justice?
7. What negative or embarrassing data can you anticipate emerging from this research? Who might be harmed (personally, professionally, financially)? What precautions have you taken?
8. Might your research lead to knowledge of sensitive matters such as illegal activities, drug/ alcohol use, or sexual behavior of participants? How do you plan to handle such information?

Comments. We must examine the impact of our research on the people whose lives we document. A classroom teacher may write field notes in order to improve her own practice. But what if her notes focus on certain members of the class ("at-risk," "Black males," "learning disabled")? These questions suggest a closer look at the power the researcher may have over students and staff (see Clay, Chapter 3, Barr Ebest, Chapter 7, and van den Berg, Chapter 8 in this volume).

Our students and colleagues are more than "subjects." Sharon Lee and Seena Kohl (see Chapter 6) suggest these distinctions:

Subject: Observed by researcher; no active participation.(Not suited to action research.)
Informant: Knowingly gives information to researcher.
Participant: More involved; perspective considered in research.
Collaborator: Fully involved in planning and interpretation.

We should also examine the impact of the study on ourselves. Becoming a researcher changes a teacher's professional status. Relationships with colleagues, administrators, and students may be threatened or enhanced. (See Minarik, Chapter 2, Clay, Chapter 3, and Hajj, Chapter 4 in this volume.)

Since I cannot be a fly on the wall in my own classroom, I must deal with my emotional and interpersonal responses as part of my data. Hammersley and Atkinson (1983) call this the principle of "reflexivity." Teacher research is engaged and committed. Its "findings" should include the relationships among participants.

V Consequences

1. Describe the possible benefits of your research--to students, teachers, or other participants; to society or to the profession.
2. Describe any risks to participants. For example, might your current students be disadvantaged for the possible benefit of future students? What steps are you taking to minimize risks?
3. Explain how you will protect the people from whom you collect data through surveys, interviews, or observations.
4. Describe how you will obtain informed consent. Do you need permission from students, parents, or both? How will you work with any students who refuse to be interviewed or to allow their materials to be quoted?
5. Are different kinds of consent needed at different stages in the project? (a) A blanket consent to participate from all students at the start of each year (with parent signatures of minors). (b) An individual consent to publish from selected students, giving you access to writing samples, videotapes, photographs, or fieldnotes that describe recognizable people.
6. Do you wish to protect the anonymity of students, teachers, parents, and other participants? If so, it is wise to use pseudonyms even in your fieldnotes. If your report is eventually published, you can also interchange physical description, grade level, gender, and so forth or develop composite rather than individual portraits. What are the gains and losses of anonymity?
7. On the other hand, instead of anonymity, it may be wiser to seek full participation and credit for students and colleagues. Research by an educator in his or her own classroom is rarely anonymous. Even if names are changed, students can be recognized in a well-written case study or classroom scene. What are the gains and losses of open acknowledgment?

Comments. These questions deal with the welfare of students and colleagues. Informed consent should be viewed as a process rather than a single gesture (see the Clayton Research Review Team, Chapter 5 in this volume). According to the AERA's Qualitative Research SIG (1993), "Informed consent is granted at the initiation of the study and codified in signed consent forms. . . . Informants may withdraw at any time, [so] informed consent is ongoing, continual negotiation" (Mathison, Ross, & Cornett, p. 3).

But how informed is "informed consent?" Traditionally, participants have been considered free from risk if the following conditions are met:

• They are first informed. They must know the general nature of the study and what is expected of them.
• They give informed consent.
• They can refuse to participate, and they can withdraw without penalty after beginning the research.
• Anonymity of persons or confidentiality of data, or both, are protected if appropriate.

On the other hand, Lou Smith argues that "field research is so different from the usual experimental approaches that many individuals, even responsible professional educators, do not understand what . . . they are getting themselves into" (1990, p. 151). He stresses the need for dialogue, moving beyond "contract" relationships to "covenants" of trust (pp. 149-150). In this volume, Lee (Chapter 6) and Smith (Chapter 9) portray the misunderstandings that can occur in the absence of this dialogue.

Qualitative researchers have long considered anonymity the norm, essential to protect "informants." Concepts from outsider social science spilled over into guidelines for teacher research (e.g., Hubbard & Power, 1993, pp. 60-61). Today some researchers, especially those working in collaborative and feminist modes, are challenging the assumptions. Why should a teacher quote published writers with credit and citations, but quote her own student writers under pseudonyms? Can anonymity conflict with intellectual property rights?

Yvonna Lincoln (1990, pp. 279-280) agrees that "privacy, confidentiality, and anonymity regulations were written under assumptions that are ill suited" to qualitative research. She suggests that colleagues, administrators, and parents participate "as full, cooperative agents," our co-researchers.
Recent models of consent forms by the National Council of Teachers of English (see P. V. Anderson, 1998, pp. 84-86) and by Hubbard and Power (1999, p. 62) give students their choice of anonymity or personal credit; a parent must approve the choice of a minor. When teacher-researchers have used this approach, most students have chosen to be named (see Barr Ebest, Chapter 7, and van den Berg, Chapter 8 in this volume).

VI Publication

1. What data will be contributed by others? Will you record student writing, oral histories, or other documents that may be considered someone's intellectual property? How have you arranged with colleagues or students for credit in your manuscript?
2. If your study is collaborative, how are you negotiating authorship and ownership? University researchers, colleagues, students, and parents may interpret their stake in the research in quite different ways. Who owns the videotape of a classroom writing group, the dialogue journal between teacher and mentor, the transcription of talk by teacher-researchers in a college seminar?
3. Who is responsible for what is said in the final report? Will other stakeholders (teacher, principal, school board) review your report in draft? Will this (a) improve your accuracy or (b) compromise your candor? Which participants (students, colleagues) might be embarrassed if they were to read your report?
4. You will inevitably gather more data than you "need." Consider why you choose some data to report to a wider audience and why some is left in your files. (On what basis do you select?) Consider the politics of the way you focus your story.
5. How will your report recognize the perspectives of participants who disagree with some of your interpretations? For example, you may revise your views; quote their objections and tell why you maintain your original view; or invite them to state alternative views in an appendix.
6. Have you decided on anonymity or on full acknowledgment of other participants in your report? Perhaps you will identify teachers but use pseudonyms for students. If you began your study with a blanket consent form, have you now requested consent to publish specific material from specific people?

Comments. Publication brings out ethical dilemmas that may lie dormant during a study. Recent discussions (P. V. Anderson, 1998) call into question the practice of relying on a general "consent to participate" when colleagues or students are to be quoted in an eventual publication. For example, students may be shown not only the samples of their work that will be published but also the passages in which they will be quoted. (This practice may lead researchers to edit such comments as "John, a low-skilled B.D. male, produced the following garbled response.")

Action research can support democratic, student-centered pedagogy when teachers periodically share their drafts with students or colleagues. Lather (1991) sees "the submission of a preliminary description of the data to the scrutiny of the researched" (p. 53) as an emancipatory approach to inquiry and a source of "face validity" (p. 67).

A growing trend in practitioner research is "multivoiced" publication, incorporating not only the perspectives but also the texts of colleagues, students, and others. Sometimes these reports take the form of a script or simulated dialogue (Harris, Lowenstein, & Scott, Chapter 11 in this volume); sometimes they weave together several authors' statements on a common topic (Zeni & Prophete, Chapter 10); sometimes they use quotations and layers of feedback (van den Berg, Chapter 8). Incorporating other voices, especially when their views differ from that of the researcher, can make a report richer, more nuanced, more authentic (McCarthy & Fishman, 1996).

VII Ethical Questions Specific to "Insider" Research

1. Does your district have a formal review procedure for research? If you are collaborating with people at a university or research institute, you may need approval from the institutional review board (IRB) in both settings.
2. Which participants at your school or college have read your research proposal? Which ones have been informed of the research orally in some detail? Which ones know little or nothing of this project? Reflect on the decisions behind your answers.
3. What do your students know of your research? Who told them? What are the risks to them or their families of their knowing (or not knowing) what you write or collect?
4. How do your school administrators view your work? Is action research under suspicion or is it mandated from the top in a drive for organizational quality control? How safe do you feel in this institutional setting pursuing this research? Would you be free to report your findings and interpretations to a wider audience?
5. Who is sponsoring this research through grants, contracts, released time, course credit, and so forth? Will you evaluate the sponsor's program, textbook, or method? Do you anticipate pressure to report what the sponsor wants to hear?
6. Does your study evaluate your own effectiveness or a method to which you are committed? How will you handle the temptation to see what you hope to see? How will you obtain other perspectives--for example, classroom observation or analysis of student work by people who do not share your assumptions?

Conclusion
Practitioner researchers who have worked through this guide may be unnerved by the sheer range of ethical concerns. It seems that three major tasks are facing us. First is to know the Federal and institutional regulations that may apply to our work. Second is to educate the people charged with enforcing those regulations so that they do not force us to mimic an inappropriate mode of inquiry. Third is to continue developing our own professional discourse about ethics.

Paul Anderson (1998) calls for an ethical standard "beyond the Federal regulation" (p. 72). Along with obtaining whatever consent is legally required, we must protect our relationships of trust with those vulnerable to exploitation, especially students.

Marian Mohr states it this way: Teacher researchers are teachers first. They respect those with whom they work, openly sharing information about their research. While they seek understanding and knowledge, they also nurture the well-being of others, both students and professional colleagues (Chapter 1 in this volume).
Whenever possible, parents, students, and colleagues should be knowingly involved in the work from the start, with time to ask their own questions and make suggestions (See the Clayton Research Review Team, Chapter 5, and Cohn & Kirkpatrick, Chapter 12).

Collaboration and communication are the best guides to preventing the ethical dilemmas of practitioner research.

Acknowledgments
An earlier version of this chapter was published in 1998 under the title "A Guide to Ethical Issues and Action Research," in Educational Action Research, vol. 6, no. 1, pp. 9-19.

The guide was originally prepared for the St. Louis Action Research Collaborative (ARC) and the Gateway Writing Project, the local site of the National Writing Project. The topics addressed reflect the suggestions and experiences of the ARC's Teacher Educators' Seminar, especially Lou Smith (Washington University), Sharon Lee (Webster University), Kathryn Mitchell-Pierce (Clayton Public Schools), Sunny Pervil (Maryville University), and Mike McGrath, Jori Martinez, and Owen van den Berg (National-Louis University). Douglas Wartzok (University of Missouri-St. Louis) also gave helpful feedback.