Chapter 5 8 Ja

Within-Case Displays: Exploring, Describing and Predicting


 Description: Making complicated things understandable by reducing them to their component parts
 Explanation: Making complicated things understandable by showing how their component parts fit together according to some rules- as Theory- Including a range of activities: providing requested information or descriptions
 Prediction: Making about the probable evaluation of case events or outcomes over the following months or years

Ladder of Analytical Abstraction

 Summarizing and packaging the data- Creating a text to work on: Reconstruction of interview tapes as written note and Trying out coding categories to find a set that fits
 Repacking and aggregating the data- Identifying themes and trends in the data overall: Searching for relationship in the data
 Developing and Testing propositions to construct and explanatory framework- Testing hypotheses and reducing the bulk of the data for analysis and Integrating the data into one explanatory framework

Display Format

 There are several formats to display the data:
– Matrices: “Crossing” of two lists, set up as rows and columns
– Networks: “Nodes” or points connected by lines

Display Type

 There are several types of Partially ordered display:
– Time-ordered: Understanding the flow and sequence of events and processes
– Role-ordered display: Understanding the role the people who live in groups and organizations

Type of Good Explanation

 Explanatory Effects Matrix is first step in the direction of answering such questions:
– Why were these outcomes achieved?
– What caused them- either generally or specifically?
 The matrix helps to understand things temporally
 It looks at outcomes or results of a process, and turn to the “what leads to what”
 It is hard to see the links between assistance types, roles and effects
 Causal Networks: Pull together independent and dependent variables and their relationships into a coherent picture
 It is useful to describe the meaning of the connections among factors

Causal Map

For the causal network, the researcher should proceed the causal map:

 Consider a specific case for which your data
 Translate the pattern codes into variables
 Rate the variable
 Draw a line between pairs of variables
 Draw a directional arrow


For predicting, the researcher should follow the procedures:

 Generating the prediction: The analyst predicted that the project would be phased out gradually in the following year
 Justifying the prediction: How do you puzzle out the “why” of a prediction?
 Analyzing the feedback: How good is the prediction? How good are the supporting and spoiling factors?

Cross-Case Displays: Exploring, Describing and Predicting

Strategies for Cross-Case Analysis

There are three strategies for Cross-Case Analysis:

 Case-oriented strategies
– A theoretical framework is used to study on case in depth
– Check whether the pattern found matches with the previous cases
 Variable-oriented strategies
– Focus on variables and their relationship
 Mixed strategies
– Combine or integrate case-oriented and variable-oriented

Working Principle

 Understand the case: Understand each particular case before proceeding to cross-case explanations
 Preserve case configurations: Must be protected the network of conditions during analysis
 Combine variable-oriented and case-oriented strategies

Methods for Cross-Case Analysis

There are several methods for Cross-Case Analysis for both of case-ordered and time-ordered:

 Meta-matrices:
– Assembling descriptive data from each of several cases in a standard formant
– Inclusion of all relevant data
 A case-ordered meta-matrix: The case are ordered according to the main variable
 A time-ordered meta-matrix: columns organized sequentially by time period
 Scatter plots
– Display data from all cases on two or more dimensions
 Scatter plots over case: Data from cases are scaled carefully, and the cases positioned in the space formed by respective “axes”
 Scatter plots over time: Scatter plots can be useful when they display similar in cases over time periods
 Case-ordered effects matrix:
– Sorts the cases by degrees major cause being studied, and shows the diverse for each case
– The focus is on out-coming dependent variable
 Case-ordered Predictor-Outcome matrix:
– Array cases on a main out-coming criterion variable, and provide data for each case
 Variable-by-Variable matrix:
– Includes two main variables in its rows and columns; specific indicators of each variable are ordered by intensity
 Causal Model:
– A network of variables with causal connections among them, drawn from multiple-case analyses
– The Principle is one of theory building
– Place causes and effects in a linear chain: Causal Chain

Reference: Qualitative Data Analysis

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License