Methods of Data Analysis in Qualitative Research
In advanced studies, a researcher may approach his topics quantitatively, qualitatively or with the use of a mixed methodology. When opting for a qualitative approach, researchers have several options in analyzing the data. The use of matrices, charts, tables and other visual displays are common tools used. With visual displays, the researcher can pare down the often abundant subjective data that has been gathered and determine what will be useful variables in his qualitative data analysis.
1 Context Chart
This display method is used to understand the context of the data found. When building thematic frames, the data included in each frame must be connected by context to be useful. Once the context chart is complete, partial analysis (partial analysis is often used to validate variables or themes) or interim analysis (interim analysis is finding an early direction or theme in the data) can be performed on the data findings. By using the context chart, the researcher shows the interrelationship of the data while keeping the research questions in mind. The researcher also strengthens the thematic framework of his study.
2 Checklist Matrix
This display method will determine whether the data is viable or useful as a variable (a variable is an object used for comparison such as 'apples' and 'oranges') in the analysis of the qualitative data. The components of the data are broken up by thematic points and placed in labeled columns, rows and point guided rubrics (e.g., 1-strong, 2-sketchy, 3-adequate) within the matrix. The thematic points are then examined for usefulness as a variable according to the numerical strength of the point-guided rubric.
3 Pattern-Coded Analysis Table
This table is created with rows labeled with themes and columns labeled by coded patterns. Pattern coding is a way to add further distinction to a variable-oriented analysis of the data. Often referred to as a cross-case analysis table, the researcher can, at a glance at the rows, render a preliminary analysis of the data collected just by noting which cell the pattern-coded data fills under certain thematic rows.
4 Decision Tree Modeling
This display method is a chart structured from one central directive. It often resembles a tree with branches. For example the central directive may be whether to buy a contract. From that directive two decision boxes are created: Pro and Con. After taking a survey, the researcher creates a branch from the pro and con boxes, allowing for a third branch for the undecided. Because the data was collected subjectively/qualitatively, the researcher will have coded the responses earlier by context to determine by pattern if they fall under pro or con. In this display the researcher will write those patterned responses in boxes resembling twigs growing from the appropriate branch to analyze the findings.
- 1 "Qualitative Data Analysis"; Matthew B. Miles & A. Michael Huberman; 1994