Action Research
Overview
Action research is a cyclical methodology that integrates research with practice improvement. Originating with Kurt Lewin, it follows iterative Plan-Act-Observe-Reflect cycles where practitioners investigate their own context, implement changes, and generate knowledge simultaneously. Participatory Action Research (PAR) extends this by positioning community members as co-researchers, emphasizing democratic knowledge production and social transformation.
When to Use
- Practitioners want to systematically improve their own professional practice
- Research must produce both knowledge and tangible change in a specific context
- Community stakeholders need to be co-researchers, not just subjects
- The problem is situated, practical, and requires iterative solution development
When NOT to Use
- When the researcher is an external observer with no stake in the practice
- When generalizable theory (not local improvement) is the sole goal
- When participants cannot commit to multiple cycles of reflection and action
- When organizational or political constraints prevent implementing changes
Assumptions
IRON LAW: Action research requires DUAL commitment — to generating
knowledge AND to improving practice. Research without action is
conventional research; action without reflection is just practice.
Both halves are non-negotiable.
Key assumptions:
- Knowledge is produced through action and reflection, not detached observation
- Practitioners are knowledgeable agents capable of researching their own practice
- Change and understanding develop iteratively through multiple cycles
- In PAR, those affected by the research have the right to participate in its design and conduct
Methodology
Step 1: Plan
Identify the problem or area for improvement collaboratively. Conduct reconnaissance (fact-finding about the current situation). Develop an action plan based on the diagnosis. In PAR, ensure stakeholders co-design the plan.
Step 2: Act
Implement the planned intervention or change. Document what was actually done (which may differ from the plan). Keep the intervention scope manageable for one cycle.
Step 3: Observe
Collect data systematically during and after the action. Use multiple methods: fieldnotes, interviews, surveys, documents, student work, video. Focus on both intended and unintended consequences of the action.
Step 4: Reflect
Analyze the data collaboratively. Evaluate what happened and why. Identify what worked, what did not, and what was surprising. Feed insights into the next cycle's planning. Document learning for knowledge generation.
Repeat the cycle — each iteration refines both the action and the understanding.
Output Format
markdown
## Action Research Report: [Context]
### Problem and Context
- Problem: [what needs to change]
- Setting: [where the practice occurs]
- Stakeholders: [who is involved as co-researchers]
- PAR elements: [how participation was structured, if applicable]
### Cycle Log
|-------|------|-------------|-----------------|-------------|
| 1 | [intended action] | [actual action] | [data summary] | [lessons learned] |
| 2 | [revised plan] | [actual action] | [data summary] | [lessons learned] |
| 3 | [revised plan] | [actual action] | [data summary] | [lessons learned] |
### Knowledge Generated
- Practical knowledge: [what works in this context and why]
- Theoretical contribution: [how findings extend understanding beyond this context]
### Practice Changes Implemented
1. [Change that was adopted and sustained]
2. [Change that was adopted and sustained]
### Validity Criteria
- Outcome validity: [did the action resolve the problem?]
- Process validity: [was the methodology sound?]
- Democratic validity: [were stakeholders meaningfully involved?]
- Catalytic validity: [did participants develop deeper understanding?]
- Dialogic validity: [was the research subjected to peer scrutiny?]
Gotchas
- A single cycle is NOT action research — the iterative spiral is essential; plan for at least 2-3 cycles
- "Participatory" in PAR means shared POWER in research decisions, not just consultation or data extraction
- Action research validity criteria differ from conventional research — use Herr and Anderson's five criteria, not internal/external validity
- The dual role of researcher-practitioner creates ethical complexity: informed consent, power dynamics, and confidentiality need explicit attention
- Do NOT treat action research as a post-hoc justification for changes already made — the research must inform the action, not merely document it
- Journal/memo writing throughout the cycles is critical for capturing the reflective dimension
References
- Lewin, K. (1946). Action research and minority problems. Journal of Social Issues, 2(4), 34-46.
- Kemmis, S., McTaggart, R., & Nixon, R. (2014). The Action Research Planner: Doing Critical Participatory Action Research. Springer.
- Herr, K., & Anderson, G. L. (2015). The Action Research Dissertation: A Guide for Students and Faculty (2nd ed.). Sage.