Module 3 – A Different Data Analysis

Erin E. Peters

April 4, 2006

Prior experiences as student

Why I wanted to try it out

Originally I coded the data I collected from teacher interviews and student focus groups into categories because it was the only way I knew how to analyze qualitative data. I have attached the codes to show how I proceeded from verbatim clips of the text to categories to illustrate this point. When I conducted the first teacher interview, I concentrated on the nature of science, but I knew after an extremely short interview that the data were thin. I conducted the next interview with the teacher’s past, present and future goals in mind so that my data would be rich. Only then did I start to uncover interesting phenomena. In coding the data for categories, I found some interesting ideas, but I also had the nagging feeling that there were stories in these data that were not being described by the codes. Some of the information that was left out in the coding scheme was teacher experiences as a science student and teacher experiences with science content during family excursions. I also felt that the teacher/student interactions were not fully explained by the categories.

What was new to me

All of the qualitative analyses I did before this project (all 4 of them) were done according to categories. I made matrices to find 2-dimensional connections and I coded using NVIVO to find overlapping categories for a cross case analysis, but none of these types of analyses involved looking at connections. These data for the inquiry project (attached) seemed to better fit a timeline than separate “bins” that characterize the topic being studied. Looking at the data to find connections rather than characteristics was the new part for me.

How I applied the method

I used the diagrammatic method described in Mason’s book on pages 169 because I realized that the codes I found were also set in the context of time. Laying out the process in a sequence gave the analysis more depth than did my original attempt at placing codes into bins. Making a visual diagram in order to describe the analysis part of a project is very appealing to me since I tend to be a visual learner. Since I wanted to look more at the connections among data, I felt that this was an appropriate method to describe my findings. I started by drawing a flowchart showing teacher and students in two interactive bubbles. I then put a second layer of bubbles containing the actions/beliefs/experiences of the teacher and the students connected to the appropriate bubble. It looked quite a mess, so my next step was to talk to someone about it. Having to explain my tangled web helps me to find out what I am REALLY trying to say. I spoke to a colleague about my web and immediately it occurred to me that I was really talking about two parallel timelines, one for the teacher and one for the student. I wanted to show how the teacher began teaching with traditional methods, because that was her only experience as a science student. Later in the first few years of her career as a teacher, she shifted her pedagogy to an inquiry method. One of the more interesting findings of this study was that at the beginning of the inquiry unit the teacher taught, the students didn’t want to switch from their usual didactic way of learning. It is clear on the timeline that the teacher shifted her stance in teaching much earlier than when the students shifted their stance toward an inquiry mode. The categories I came up with previously did not illicit any of these connections. The fact that I diagrammed the data in a timeline illuminated this piece of information that was not apparent in the coding.

The process I used to make the diagram began with looking at the data in a chronological order. I went back to the data and re-told the teacher’s and students’ stories according to a chronological order. The order that I used for the teacher data was 1) experiences as a science student, 2) science experiences as a child with the family, 3) early teacher experiences, 4) shift away from teacher as sole authority, 5) current science teaching, 6) scaffolding for students, 7) negotiating with students. The order I used for the student data was 1) experiences early in this class, 2) difficulties with the inquiry structure of the class and 3) coming to terms with the inquiry structure of the class. When I read the data in this order, I plotted them on a timeline. After I plotted the parallel timelines I saw how the teacher scaffolded information for the students and how they both negotiated three common working ideas in order to make sense of an open ended process. I also saw one place where the teachers and the students had oppositely oriented and simultaneous reactions to the process.

What I learned from this

The most important thing that I learned from this exercise was that there may be deeper messages in data than just comments from the participants. When I coded the data, I could only see the obvious things that were mentioned during interviews and focus groups. When I organized the data in a different way to make a meaningful diagram, I was able to see below the surface inferences. Having said that, I am also aware that there are many, many ways to interpret data and that trying to see too much into the data can be a hazard. However, now I can be better equipped to perform a validity check by looking at possible alternate inferences from the data. If I can look at my data through various analytical methods, I can find alternate inferences and decide on the most valid conclusions.