Friday, July 9, 2010

Capabilities of Computer Software in Qualitative Research

"Computers used in qualitative research can be useful in conducting mechanical tasks such as searching and arranging descriptive data, but they cannot perform the conceptual tasks."

Some common tasks associated with qualitative research that have been programmed for computers are:

1.  Coding - requires an organized set of categories based on a specified rationale.  Coding "themes" or categories commonly are drawn out of the data. 

2. Lexical Searching - looking for specific words, phrases, combinations, or strings of words.  Reveals the dominant terminology used to describe the phenomenon under study.
 
3. Hypertext - is a flexible, database management system where segments of text in the research data are joined together by electronic links.  The data may be explored in various side trails.


Various software packages include:
1. The Ethnograph 

2. NUD*IST  (Non-numerical unstructured data - indexing, searching, and theorizing)

3. NVIVO (an upgrade of NUD*IST)

Reliability and Validity of Qualitative Research

"Validity of qualitative research for the most part is established on a logical basis, and providing an argument for validity requires well-documented research and a comprehensive description."

- traditional concepts of reliability and validity may cause qualitative researchers some difficulty because qualitative research occurs in natural settings which are extremely difficult to replicate.

- internal reliability is consistency in the research process which may involve observation by multiple observers or the use of videotapes.

- typically there will be some disagreement among observers, and one way to obtain a measure of agreement is to complete the ratio of agreements to agreements plus disagreements...this is called concordance.

- interval validity relies on the logical analysis of the results.

- since research is conducted in natural settings  with complex phenomena, controlling variables is difficult.  Thus, verifying the results and conclusions in qualitative research is enhanced when using two or more sources or perspectives.

- external validity is more concerned with the comparability and translatability of the research (not with broad generalizations of results).

- comparability refers to the extent to which adequate theoretical constructs and research procedures are used so that other researchers can understand the results.

Funnels and Stuff...

TASK 1:  Your group has been given a blank copy of both the funnel and modified analytic induction approaches, as well as two ziplock bags that contain the stages of each approach.  Your task is to correctly order the stages in each model.

Funnel Approach Template:




Modified Analytic Induction Approach Template:



Funnel Approach Terms:




Modified Analytic Induction Approach Terms:




TASK 2:  Now that you have the correct order of each stage for each model, you have been given two more ziplock bags which contain specific information regarding research being conducted on a constructivist approach to instruction in elementary school reading and mathematics, grades 2-4.  Your second task is to correctly place each descriptive card under the correct heading for each model.

Funnel Approach Explanations:




Modified Analytic Induction Approach Explanations:




TASK 3:  Choose one of the discussed models and apply it to the research question you are writing your own proposal on.

Types of Design in Qualitative Research

Research design in qualitative research is generally less structured and more flexible than research design in quantitative research.  Research Designs differ because of the context, purpose, and nature of the research.


Qualitative Research

Because research design is a process, it is useful to think about research design in terms of activities or tasks conducted during the research (p. 211).  Qualitative research may be approached from two perspectives:

1. FUNNEL APPROACH - Begins with general research questions that initiate the study, then on the basis of results from initial data collection the study conditions of the study are identified more specifically.  This process may be repeated to further focus the study until the conclusions are concentrated on a specific, or limited number, of components.  

2. MODIFIED ANALYTIC INDUCTION APPROACH - Begins with more specific research problems or questions and then attempts to cover all cases of the phenomenon under study to arrive at a comprehensive, descriptive model.  There is a reformulation of the model to accommodate data which does not fit the model.  This process of data collection and reformulation is continued until a satisfactory, universal explanation is obtained.


Interactive Activity #2:  Using the model you are assigned, follow the model process to design a qualitative research study on the nature of instruction using a constructivist approach to teach reading and math in grades 2-4.

Figure 8.1 Components of Research Design in Qualitative Research

The diagram below appears on page 208.  It is a representation of the necessary components of research according to Wiersma.  Wiersma notes that these components are both highly integrated and interdependent.






Interactive Activity: Fill in the Blanks

For this activity you will be given a card or two with a word or phrase on it.  Your job is to study the card and determine where your word/phrase fits on the overhead diagram.

Coding

"Data analysis in qualitative research is a process of categorization, description and synthesis.  Data reduction is necessary for the description and interpretation of the phenomenon under study."


Coding:

  • coding is the process of organizing data and obtaining data reduction
  • it is the process by which researchers "see what they have in the data"
  • the researcher needs to determine an appropriate category system to organize the data
  • any rational system of codes that benefit the research can be used
  • setting or context codes are usually very useful
  • the subject's perception codes are also commonly used.
  • process codes, which focus on the sequence of events and how change occurs, are also useful

Thursday, July 8, 2010

Components of Research Design

"The working design is a preliminary plan for getting the research under way."


Working Hypothesis:

  • Qualitative research uses inductive inquiry
  • Data collection is undertaken without preconceived theories or hypotheses
  • Researchers are influenced by their backgrounds
  • Questions, hypotheses and foreshadowed problems may be reviewed, deleted, or extended as the data collection and analysis proceed

    Data Collection:

    • the researcher must gain access to do the research
    • data collection can be interactive or non-interactive
    • most common methods include observation, interviews, collection and reviews of related documents, taking specimen records, and taking oral histories.

    Data Analysis and Interpretation:

    • Analysis and Interpretation begins shortly after data collection begins
    • it continues as data is collected
    • concurrent collection and analysis allows the researcher to check working hypotheses and unanticipated results
    • qualitative data analysis requires organizing information and data reduction
    • early data might suggest a theory
    • more data might support, dis-confirm or extend the hypothesis or theory
    • The data may suggest categories for characterizing information. Coding is a process where the researcher decides on one or more categories for organizing data. 
    • Qualitative research is a process of successive approximations toward an accurate description and interpretation of the phenomenon.

      Wednesday, July 7, 2010

      The Epistemology of Qualitative Research

      "The epistemology of qualitative research provides the underpinnings for how qualitative research is conducted-how data are collected and analyzed and how conclusions are reached."