Action Research¶
Table of contents
- The "Action-Reflection" Spiral Cycle of Action Research
- How to Organize an Action Research Study
- Application Cases
- Advantages and Challenges of Action Research
- Action Research Cycle Planning Template
- Example: Reducing Handover Errors on a Hospital Ward
- Judging Quality in Action Research
- Common Mistakes
- Frequently Asked Questions
- Extensions and Connections

In traditional research, researchers typically play the role of objective, detached "observers," while research subjects are passively "studied." Action Research completely breaks down this barrier. It is a cyclical inquiry process that tightly integrates "research" with "practice," aiming to solve practical problems and drive change. Its core idea is that knowledge should not merely be "discovered" and left on the shelf, but should be "created" and "applied" in the action of solving real-world problems.
Action research is not a "prescription" given by external experts to practitioners, but a systematic process of diagnosis, reflection, action, and evaluation of one's own work context (such as a classroom, a community, an organization), carried out by practitioners themselves (often in collaboration with researchers). The core question it answers is: "How can we improve our current work/situation?" Therefore, action research is characterized by its strong contextuality, participation, collaboration, and cyclical nature. It is both a process of understanding the world and, more importantly, a process of transforming the world.
Key takeaways
- Research with people, not on them: participants help define the problem and judge results.
- A spiral of cycles: plan → act → observe → reflect → revised plan.
- Two goals at once: improve a real situation and generate knowledge about it.
- Common in education, healthcare, community development and organizational change.
- Rigor comes from systematic data, documented reflection and multiple perspectives.
The "Action-Reflection" Spiral Cycle of Action Research¶
The soul of action research lies in its continuously iterative spiral process. Each cycle consists of four closely linked stages, and this cycle repeats constantly, with each iteration deepening the practitioner's understanding of the problem and making improvement actions more effective.
The most classic action research spiral model was proposed by Kurt Lewin and has been continuously developed by later generations:

How to Organize an Action Research Study¶
-
Identify a "Practical Problem" Worth Addressing Action research begins with practitioners' confusion, dissatisfaction, or desire for improvement in their work. This problem must be specific, real, and one that the practitioner is motivated to solve. For example, a teacher finds that "students' participation in class discussions is low."
-
Form a Collaborative Team and Conduct Preliminary Diagnosis Invite all stakeholders related to the problem (e.g., other teachers, student representatives, researchers) to form a collaborative group. The team jointly conducts an in-depth diagnosis of the problem and analyzes its possible causes.
-
First Round of Cycle: Plan, Act, Observe, Reflect
- Plan: The team decides to try a new teaching method – "group debate competition" – and develops a detailed implementation plan.
- Act: Over the next month, the teacher implements four "group debate competitions" in her class.
- Observe: During this period, she collects data on student participation, quality of发言, and feelings through classroom recordings, student interviews, and her own teaching log.
- Reflect: The team jointly analyzes the data and finds that although overall participation significantly increased, some introverted students still spoke little. At the same time, the debate competition took up too much teaching time.
-
Enter the Next Round of Cycle
- Re-Plan: Based on the previous reflection, the team revises the plan. The new plan is adjusted to "online group discussions before class + representative summaries in class," and different roles such as "online scribe" are set up for introverted students.
- Act, Observe, Reflect again... This spiral process will continue until the problem is satisfactorily resolved, or the team has formed a deep understanding of the problem.
Application Cases¶
Case 1: Teacher Professional Development
- Scenario: A high school math teacher wants to improve students' understanding of the abstract concept of "functions."
- Application: She collaborates with a math education researcher from a university to initiate an action research project. They jointly design a teaching plan based on real-life situations (e.g., mobile phone bill packages) (Plan) and implement it in the classroom (Act). By analyzing students' homework, test scores, and class discussions (Observe), they find that students' understanding of functions has indeed deepened, but their ability to perform symbolic operations has not improved much (Reflect). In the next cycle, they adjust the plan, adding targeted exercises for symbolic operations.
Case 2: Community Development and Empowerment
- Scenario: Residents of an old community generally complain about the lack of vitality in public spaces and indifferent neighborly relations.
- Application: Several community workers and resident representatives launch a "community garden" action research project. They jointly plan a wasteland (Plan) and organize residents to transform and plant it together (Act). During the process, they observe a significant increase in residents' participation and neighborly interaction through interviews and activity records (Observe). At the project summary meeting, everyone reflects on the successful experiences and decides to promote this model to other public space renovations in the community (Re-plan).
Case 3: Organizational Process Improvement
- Scenario: A software company's development team finds that their product release process is too lengthy, often leading to delays.
- Application: The team decides to use action research to optimize the process. They first draw the existing process flow chart and identify bottlenecks (Plan). Then, they decide to pilot "continuous integration" in the next release cycle (Act). By tracking data such as code submission frequency, bug count, and release time (Observe), they find that the new method significantly shortens the release cycle (Reflect). Thus, the team decides to promote this successful practice throughout the company.
Advantages and Challenges of Action Research¶
Core Advantages
- Directly Addresses Real Problems: Research directly stems from practice, aiming to solve real, specific problems encountered in practice.
- Empowers Practitioners: Transforms practitioners from passive "research subjects" into active "researchers," greatly enhancing their professional reflection abilities and autonomous problem-solving skills.
- Bridge Between Theory and Practice: Tests theories in action and develops theories in reflection, effectively connecting theoretical knowledge with practical wisdom.
- Promotes Sustainable Change: Since changes are driven from within by internal personnel, they are more easily accepted and more sustainable.
Potential Challenges
- Rigor and Objectivity: Since the researcher is also a participant, maintaining the systematic nature of the research process and the objectivity of the analysis is an ongoing challenge. Detailed process recording and team collaboration are key to ensuring rigor.
- Time and Effort Investment: Action research requires practitioners to invest extra time and effort in learning, reflection, and discussion outside of their busy daily work.
- Generalizability of Conclusions: The conclusions of action research are usually highly contextualized, and their purpose is not to generate universal theories but to improve specific practices. Therefore, its research results are difficult to directly generalize to other contexts.
Action Research Cycle Planning Template¶
| Element | Cycle 1 | Cycle 2 |
|---|---|---|
| Problem / question | ||
| Planned change | ||
| Who is involved | ||
| Data to collect (before, during, after) | ||
| Timeframe | ||
| What happened (observations) | ||
| Reflection: what worked, what didn't, why | ||
| Decision for next cycle |
Example: Reducing Handover Errors on a Hospital Ward¶
- Plan: Nurses and researchers noticed missing information at shift handovers. They agreed to try a standard handover checklist.
- Act: The checklist was used on one ward for four weeks.
- Observe: Researchers sat in on handovers, counted omissions and interviewed nurses.
- Reflect: Omissions fell, but handovers took longer and night staff found the checklist too long.
- Revised plan (cycle 2): The team shortened the checklist to the five most-missed items and tested it again.
Judging Quality in Action Research¶
| Criterion | Question |
|---|---|
| Outcome validity | Did the action actually resolve or improve the problem? |
| Process validity | Was data collected and analyzed systematically? |
| Democratic validity | Were the perspectives of all stakeholders included? |
| Catalytic validity | Did participants gain new understanding and capacity to act? |
| Dialogic validity | Have findings been tested through discussion with peers? |
Common Mistakes¶
- Calling any improvement project "action research". Without systematic data collection and documented reflection, it's an improvement initiative, not research.
- Researchers deciding everything. If participants aren't involved in defining problems and judging results, the participatory core is lost.
- Only one cycle. Learning comes from iteration; the first cycle rarely gets it right.
- Weak documentation. Without records of what was tried and why, others can't learn from or trust the findings.
- Ignoring power dynamics. Hierarchies affect what people feel safe saying; create space for honest input.
Frequently Asked Questions¶
Who developed action research?
Social psychologist Kurt Lewin coined the term in the 1940s and described the cycle of planning, action and fact-finding about the results.
How is action research different from traditional research?
Traditional research usually studies a situation from the outside without changing it. Action research deliberately intervenes, studies the effects and involves participants as co-researchers.
What is participatory action research (PAR)?
A form of action research that emphasizes equal partnership with communities, especially marginalized ones, in every stage of the research, with the aim of social change.
Can action research be used in business?
Yes. Organizational change, process improvement and leadership development projects often follow action research cycles, similar in spirit to PDCA and Kaizen.
Extensions and Connections¶
- Qualitative Research: Action research extensively uses qualitative methods such as interviews and observations in data collection.
- Critical Theory: Some schools of action research (e.g., critical action research) have a strong social critique, aiming to reveal and challenge unequal power structures and promote social liberation.
- Lean and Agile: In enterprise management, the PDCA (Plan-Do-Check-Act) cycle in Lean thinking and the iterative thinking in Agile development are highly consistent with the spiral cycle concept of action research.
Reference: Kurt Lewin, a Gestalt psychologist, is widely considered the "father of action research." The works of Stephen Kemmis and Robin McTaggart have significantly developed the theoretical and practical models of action research. In the field of education, John Elliott and Lawrence Stenhouse are key figures in promoting the "teacher as researcher" movement.