The use of a variety of methods, or method triangulation, enhances the validity of action research by allowing both the convergence and divergence of data. Where data from a number of sources converges, the reliability of that data is strengthened. On the other hand, the validity of findings can also be enhanced in qualitative research when findings gathered by different means present contrasting or contradictory understandings (Ratcliff 1999). This contrast may result from the dynamic nature of the naturalistic research setting, whereby the conditions in which data are collected are never the same. Alternatively, contradiction in findings may result from the fact that different methods measure different perspectives. Whatever the cause, it is vital to report discrepancies, or negative cases, in order to support the non-judgemental nature of the research process. In addition, especially in action research, divergent cases can be an important consideration in determining the direction of future cycles of action.

Examples

·     Participant Self-reporting

- Journals

- Descriptive Surveys

- Key-informant Interviews

- Focus Group Interviews

·     Direct Observation

·     Field Notes

Participant Self-reporting

Journals

Providing participants with the opportunity for self-expression through the use of a diary or journal has a number of advantages. Firstly, the ‘private’ nature of a journal can facilitate an honesty in response that may not be shared in an interview. Secondly, journaling provides an opportunity for reflection and the possible recall of information or feelings that might not come immediately to mind in the context of a survey or interview. Finally, journal reflections provide for active participation in the research process by allowing participants to influence the future direction of the research. This empowerment of participants may lead to an increase in their efforts to own the research and to make it happen.

Surveys

The descriptive survey is another self-reporting method employed to gather data. This method has the advantage of being easy and relatively straightforward to administer, and is an appropriate method to provide a snapshot of particular phenomena. The use of a method such as this enables a broader view of phenomena to be obtained from a much larger sample than is used with a qualitative method such as the focus group interview. As the name suggests, the use of descriptive surveys is more about gathering data that can be analysed qualitatively, through verbal data and descriptive statistics, rather than quantitatively through formal statistical analysis (Tanner in Williamson 2002, p.91).

In using self-reporting data collection methods such as journals and descriptive surveys, the limitations of such methods must be acknowledged. Firstly, it is recognised that the data gathered ‘will vary considerably with students’ abilities to accurately assess their own cognitions, behaviours and affective responses’ (Assor & Connell in Chapman 2003).   Secondly, it is acknowledged that, where the research is conducted in a natural setting, it is nigh on impossible to identify cause-effect relationships, because of the possibility of rival explanations (Turner in Williamson 2002, p. 93.) Therefore, the analysis of data gathered through self-reports should not be viewed in isolation, but rather in conjunction with data from other sources in order to create a holistic picture of the phenomenon being explored.

Key-informant Interviews

Key-informant interviews are in-depth interviews with individuals who are knowledgeable about the research setting and the project’s participants. Such interviews can take the form of deep conversations in which predetermined topics are explored in a flexible and often indirect manner (Tacchi, Slater & Hearn 2003). Rather than attempting to gain hard facts, the aim is to uncover the key-informants’ perceptions and understandings of the phenomenon being explored.

Focus Group Interviews

According to Krueger (in Williamson 2002, p.253), focus groups are a particularly appropriate method to gather data ‘when the goal is to explain how people regard an experience, idea or event’. They allow participants the opportunity to express their feelings and opinions in ‘their own terms and frameworks of understanding’ (Williamson 2002, p.252). Other advantages of focus groups cited by Williamson include: the ability to produce rich data focused on a topic determined by the researcher; the opportunity for the clarification of participant responses; and the ability of participants to react to and develop the responses of others.

To avoid the potential pitfalls of focus groups, such as the influence of peer pressure on participant responses, or the introduction of bias by the interviewer, the data gathered using this method should be considered in conjunction with data collected using other methods.

Direct Observation

Given the subjectivity of participant self-reporting, direct observations are a useful method to enhance the rigour of action research by confirming self-reported data (Chapman 2003). In this project, direct observations in the classroom were primarily focused on physical and affective indicators of engagement. Assessment of students’ engagement was considered in terms of their active, passive or distracted involvement in tasks that focused upon reading, writing, discussion, groupwork and performance. Observations were made using ‘whole-interval sampling’ (Chapman 2003) which involved observing students for a short period, rather than for a ‘moment’. As Chapman indicates, whilst ‘this procedure will produce relatively conservative estimates of student engagement rates, it is also likely to be more sensitive to variations in the consistency and persistence of students’ behaviour’.

Field Notes

Notes taken during classroom observations can take the form of written observations, as well as video and still photos. The old adage that ‘a picture tells a thousand words’ is particularly relevant for action research. According to Ratcliff (1999), the ‘tangible, concrete nature of pictures… produces a more holistic view of events and situations’, and provides a re-examinable source of data that can be used to triangulate written data.   The capturing of facial expressions, and spatial relationships and interactions, for example, can provide useful evidence.

Whilst the details of an image present a situation as it is at the time of capture, it must be recognised that photography can also be selective in its field of vision. For example, there may be a tendency to photograph evidence that supports the researcher’s assumptions or hypotheses. Similarly, the act of taking photos can create a distraction in the research environment that may influence the participants’ behaviour. In an attempt to cater for these limitations, numerous photos can taken across time and space in an attempt to provide as wide a field of vision as possible and to desensitise participants to the act of photo-taking.

References

Chapman E 2003, ‘Alternative approaches to assessing student engagement rates’, Practical Assessment, Research & Evaluation, vol. 8, no. 13, viewed 17 September 2005,   <http://pareonline.net/getvn.asp?v=8&n=13>

Tacchi J, Slater D & Hearn G 2003, Ethnographic Action Research, viewed 19 January 2006, <http://unescodelhi.nic.in/publications/ear.pdf>

 Williamson, K 2002, Research methods for students, academics and professionals, 2^nd edn, Centre for Information Studies, Charles Sturt University, Wagga Wagga, NSW.

Looking to do a survey?

Zoomerang  allows you to set up a survey via a free or subscription service. The free option allows you to create a survey for up to 100 people with up to 30 questions.  When the surveys are completed, Zoomerang will analyse the results and present them in a variety of formats (these can be downloaded).  These results will be available for 10 days. If you take out a subscription, the raw data and the results are available for download  (plus there are other enhanced functions available).

If you don't need to conduct more than 100 surveys and you don't have more than 10 questions, it looks like a more viable option than Zoomerang is Survey Monkey.  If you need to conduct more than 100 surveys or include more questions, you can subscribe on a monthly basis. I'm sure Margaret would be happy to tell you more if you're interested.

By the way, at Survey Monkey's Help Centre you'll find a document that might help you in designing your survey questions.