©©  2019 Center for Collaborative Action Research  

" If we could change ourselves, the tendencies in the world would also change--" Mahatma Gandhi

  • Action Research Tutorials Videos
  • Facebook Social Icon

Tutorial 8:  Analyzing Data - Evidence

Tutorial 8: Activities

 

Organize your data into themes or topics that relate to your research question and to the outcomes that listed in the logic model. For example, if your action was to help foster a community of practice orientation at your workplace, in your logic model you might have predicted the following near outcomes: higher productivity, less duplication of services, and a more cohesive and creative working group. So for data analysis, you might start by using these categories. Then you could look at the data you collected and arrange it under each of these categories. Your approach will depend on the type of data.  See the video and resources from this tutorial for other ideas for organizing your data.

 

A. Explore- Organize Your Data Into Your Storyline:

 

 

  • Descriptive data (numerical or quantitative): Where you have data that can
  • be counted, compute the numbers. Size matters in reporting descriptive data. Always tell your reader how large your data set is. N =15 is the shorthand for the number of respondents. Percents are only meaningful information when you know the size of the whole. If you say 50% of the students, this could mean 10 or 200 depending on who participated. You have to use percents when you are comparing groups of different sizes. So if you are working with two teams and one has 20 people and the other 10, and you only report that 10 in each group responded yes, one might assume this response was the same for both groups. But in fact, 50% of group 1 and 100% of Group 2 responded yes. If you only have one group and there are less than 10 respondents, a number is often a better representation of the value than a percent.
  • Narrative or textual data (qualitative): Analysis of student blogs or opened ended questions often involves coding. The codes can come from the data or from your research questions. Suppose you asked students to reflect on what they learned from a project you designed. Read through the responses and come up with a plan for coding. Often this entails writing notes in the margins and then looking for patterns. If you start to see something that many of the students are saying you give that a code. For example, students might mention some of the concepts they learned, some of the technology they mastered, or some changes in attitude or identity. You could then code each response to see if it mentioned one or more of these outcomes. If your goal was to increase content knowledge, then you might only code comments about knowledge but you might have five level of knowledge. In this case, each response could be coded for evidence of the depth of understanding. And if you have more data than you can code, you can "sample" by taking a subset of responses. Sometimes a random sample works, and other times a random choice within categories is better. For example, you could group your class by grades received on the last report card and randomly sample blogs of three students from each of the grade groupings.

  • Media Data: If you took pictures you can study the pictures for evidence. For example if you took pictures of your class every 10 minutes during a project or have a video that you can stop at regular intervals, you could count how many students look like they are fully engage in teamwork and how many appear to be off task. With an audio of a meeting, you might focus only on question asking-- who asks questions, how often and what type of question.

 

 

 

C. Visualizing - Display your Data to Tell Your Story:

Once you have counted your data you will be ready to create charts, tables, graphs or other ways to help your audience see at a glance what you had to spend hours finding. Whatever tool you have used, share it with someone who does not know what you have done and ask them what they see. This is the communication test, does it convey what you intended.

B. Analyze - Examine your Data to Find Your Story 

 
 

D. Writing -  Your Cycle 1 Report


 

  • CYCLE RESEARCH QUESTION: This question needs to contain two very important parts. The first part clearly states what you did. And the second part shares your best guess at the outcome(s) you anticipate. Your action research is a design experiment. You are designing with an eye towards a deeper understanding of the consequences of actions.

  • ACTION TAKEN: Describe what you did in enough detail that we understand the action that was taken.

  • EVIDENCE USED TO EVALUATE THE ACTION: Describe the data that you collected to give you direct or indirect evidence of what happened.

  • EVALUATION: Share the summary of the data analysis process from these activities

  • REFLECTION: (to be added after the cycle is complete) see next tutorial

 
 

E.  Sharing - Discuss Your Ideas with Action Researchers

 

Your learning circles or critical friend discussions are very important. Talk is the central vehicle for sharing, analyzing and evaluating actions that define your practice. It is how we learn. Hopefully, you are working with a group of people engaged in action research either through a university or a more informal learning circle (see onlinelearningcircles.org). If you are working alone, you might want to use the forum on this site to share ideas. Here are some ideas to support these discussions:

  • 1) Revisit the description of your practice and logic model--provide an overview of what you are doing and why. This will be helpful as you develop your action research portfolio

  • 2) Attention to evidence--Bring your data and analysis plan to the forum and look carefully at the plans that others bring. Concentrate on developing alternative interpretations and other possibilities. We are often blinded by what we think. We depend on each other to see without the blinders. That is one of the fundamental reasons for learning circles, to share an alternate way of thinking. Do NOT be afraid to challenge data as this will help action researchers to see past what they think.


In a discussion, think about the type of feedback that builds knowledge. Some knowledge building ideas included listening carefully to what others say and ask them to clarify.  When you are asked a question, ask the person why they asked you that question. Challenge each other to think about what evidence can be used to support claims like "it went so well" or "I was so happy with the outcome."