Last reviewed 2 July 2012

What is action research and how can you use it in your school? Luci Boella discusses the practicalities of using this type of research and analyses its usefulness.

Using action research

Being reflective is central to becoming a successful practitioner. Furthermore, teachers are increasingly being held accountable for students’ levels of achievement and exam results so, in return, it is becoming increasingly important for teachers to engage in a systematic method of improving or justifying practice. Action research is undertaken within the school setting for the purpose of gaining greater understanding of teaching practices or as a tool to improve practice. It provides a systematic, straightforward method of assessing the requirements of learners, of collecting and analysing data pertaining to those requirements, and of documenting the outcomes of the investigation. In turn, this allows informed decisions to be made on teaching practices or learning environments and identifies the benefits of particular methods.

There are different types of action research depending on the scope of the research and the individuals undertaking it. Most commonly, it involves one teacher who identifies an area for improvement, such as with a particular class or more generally in their personal practice or with their use of materials. The research that ensues is then on a small scale and either involves collecting observational data (eg questionnaires completed by students) or collecting performance data from students to assess where and when learning is superior. However, action research can also be employed on a larger scale. It can range from a collaboration between two or more teachers that focuses on issues at the departmental level or more global issues that permeate the whole school. It can even be used at a district level where education authorities use feedback from schools to inform local policies. Whatever the scale of the research, it can be a highly rewarding and informative process for the individuals involved. In addition, feedback from students who participate indicates they also enjoy the process and feel empowered by participating in research that can impact on how they are taught.


The first point to consider when undertaking action research is its purpose. Is it a “fishing” exercise to expose opinions (of students or staff) or the strengths and weaknesses of a particular practice, or is there a clear and definite question that the research needs to answer? If the purpose of the research is not clearly defined, it might be necessary to start with an exploratory investigation and then develop a more clearly defined research question afterwards. For example, a teacher may want to examine students’ preferences about different desk arrangements. Following student feedback, the teacher may decide to trial certain arrangements to assess if class behaviour and learning is superior in the students’ preferred layout.

Once the purpose of the research has been decided, the next step is to establish if there is any previous research that can either aid in the planning of the new research or inform more generally on the topic under investigation. Some schools and universities will have extensive access to research articles through databases. However, a thorough Internet search for scholarly articles is often sufficient and as most action research is not published formally, a detailed and lengthy literature review (as would be expected in academia) may neither be practical nor possible. Even so, it is still important to assess if similar research has been carried out previously as, if nothing else, it may allow the researcher to avoid repeating the same mistakes made by others.

In designing and carrying out the research, there are several forms of data that can be easily collected using the following methods.

  • Interview: either structured with specific questions or open, where the direction of conversation is led by the participant.

  • Questionnaire: again, these can either be open, lengthy responses or closed, pre-determined multiple choice questions.

  • Performance: for example, exam results.

Interviews can provide rich and detailed data but can also be time consuming and difficult to administer and interpret without bias. Questionnaires, whilst they lose some of the richness of data, allow for large samples to be tested quickly and without bias. Closed questions with limited responses also allow for easy data analysis but questionnaires should always be piloted to ensure questions are suitably worded for their target population and elicit the required responses. Finally, performance data is easy to interpret and analyse but cannot capture opinions so is only suitable for certain research questions.

Ethical considerations

If the research is to be considered ethically sound, participants should be informed of the task prior to participating and given the opportunity to opt out. Common practice in more formal research settings is to provide participants with an information sheet describing the purpose of the research and the task they will be asked to complete, and informing them of their right to opt out. In a less formal research setting, it may be sufficient to give this information verbally. When pupils are involved, it is also necessary to consider if parental consent is required. Generally speaking, however, so long as the research does not ask a pupil to do anything over and above what would be expected of them in a normal classroom setting, parental permission is not considered necessary.

Analysing data and drawing conclusions

Determining if data are statistically significant or not is normally the sole purpose and the holy grail of social science research. However, where samples are small, statistical analysis may be meaningless as statistical tests are not designed to accurately detect significance in small samples. In such cases, it may be sufficient to ascertain the most common responses to closed questions in questionnaires or calculate average test scores/improvements for performance data. Data collected from interviews and open questionnaires can be analysed for recurring themes and particularly important details. Action research that is robust enough in its design and that has statistically significant results could be submitted to a suitable peer-reviewed journal for publishing. However, the purpose of most action research is to inform on individual or departmental practice and encourage personal development. Therefore, writing a formal report of the research may serve little purpose. Instead, the results should be used to make informed changes to practice or to inform departmental or school improvement plans. Additionally, or alternatively, the research may highlight future directions of research.

The limitations

Action research is a systematic and valid tool in helping identify areas for personal improvement and helping inform school-wide improvement plans. However, its impact is limited to just that. The research, because it is in the main not stringently planned or carried out by trained research professionals, is not often considered robust enough to be recognised outside of the community in which it was carried out. Furthermore, where research is not stringently planned, it leaves it open to alternative explanations and conclusions from those drawn by the researcher. Researcher bias is one influencing factor that it is particularly difficult to eliminate from research of this nature. It is difficult for a teacher who has preconceptions about his or her own practice or about a specific teaching method to undertake research that is entirely objective. In addition, if the pupils are informed as to the purposes of the research, they may also be biased in their responses, thus invalidating the data and conclusions.


Action research is a rewarding and powerful tool for making informed decisions about educational practice for individual teachers and for school-wide policies and improvement plans. The informal research can be carried out by any individual seeking personal improvement or informed reflection and does not necessarily require lengthy literature reviews or detailed and complex data analysis. However, action research is often carried out by individuals who have preconceptions of desired outcomes. The results, therefore, can be biased with conclusions that cannot be generalised outside of the setting in which they were undertaken or to practitioners other than those that undertook them. That said, taken as it should be — as a tool for personal development or development of the school community in which it was undertaken — action research can serve to bridge the gap between formal educational research and educational practice.