Action Research

Ernest T. Stringer. (1999). (2nd Ed.). Thousand Oaks, London, and New Delhi: Sage.

Chapter 1. Research in Professional and Public Life.

Similar to Kurt Lewin's model of action research, Ernest Stringer calls attention to action research as a decentralized small-scale format of research that would benefit practitioners' daily work in working with community groups. His model consists of a look-think-act routine. There are three basic assumptions in his model: a researcher's involvement, research subjects' participation in the research processes, and direct benefit to all participants. Thus, community-based research takes on action research. Built on Denzin and Lincoln's (1994) argument, Stringer defines as community-based research "enacting local, action-oriented approaches to inquiry, applying small-scale theorizing to specific problems in specific situations" (p. 10). Characterizing such research as stakeholders' participation, he specifies stages for the participation: data collection, analysis, theorizing, action, and evaluation. Here Stringer describes researchers as research facilitators with less directive power.

Chapter 5. Think: Interpreting and Analyzing.

Built on Denzin's (1997, Interpretive Ethnography) call for performance--"texts waiting to be staged" (as cited on the page 109)--Stringer contends that predominant formalist texts for presentation of research findings may not only silence the voices of people but also limit evocative responses from audiences. He calls particular attention to the movement to performance, such as drama, role play, dance, song, poem, etc., in action research so that people can express their feelings without intrusion of the formality.

Chapter 8. Formal Reports, Theses, and Dissertations (pp. 165-186)

This chapter has useful information about how to write up research findings in formal/informal ways. [GOOD]

In particular, the section of methodology is excellent. Stringer divides it into four subsections.

(1) Methodological Assumptions or a Philosophical Rationale: he suggests a more extended treatment than usual as PAR is not widely accepted yet. Identify research paradigm, detailed purposes, processes, outcomes of research. Cite sources. In principle, answer these questions, "why have we used this approach to research? Is it a valid and rigorous research process? Clarify the nature of the research process by comparing it with survey research, or more conventional styles.

(2) Research Processes: details, who was involved, what information collected, how analyzed. Include position of the researcher and participants. Data collection technique may include INTERVIEWS, OBSERVATIONS, DOCUMENTS (MEDIA, ARTIFACTS), RECORDING INFORMATION (VIDEOTAPES, PHOTOS...). The part of analysis should provide how research participants analyzed information. Stringer introduces the term BRAKETING to refer to the process of analyzing information in interpretive research to reduce it to its most significant or key, elements.

(3) Rigor: Unlike traditional criteria for evaluating rigor, such as objectivity, reliability, validity and generalizability, Stringer proposes CREDIBILITY, TRANSFERABILITY, AND DEPENDABILITY AND CONFIRMABILITY in order to establish the trustworthiness of PAR.

He asserts that credibility is established by prolonged engagement with participants, triangulation of information from multiple sources, member checking to verify the accuracy, and peer debriefing for reflection on the process.

By Transferability Stringer means applying research findings to other contexts. He argues that research methods should provide thick description so that audiences can identify similarities of the research setting with other contexts.

Dependability and confirmability are obtained through detailed description of collecting and analyzing processes and references to the raw data.

[I think that my research project can have certain credibility to begin with. I should work in transferability and the other. Should I keep the raw data? Final video should be sufficient as raw data if we publish a report in a formal way?]

Stringer also suggests having a section of limitations, not for an apology, but for an explanation about why traditional criteria such as objectivity, reliability, etc. are not appropriate.

(4) Ethical issues
guard against unwarranted intrusion into private life, maintaining privacy, establishing ownership, describe how research processed were enacted in ways sensitive to the cultural values and protocols of research participants.

Chapter 9. Understanding Action Research: The Theory Behind the Practice (187-215)