Action Research
Supplementary Materials for LSI 580 Evaluation and Research
Department of Library Science and Instructional Technology
Southern Connecticut State University, New Haven CT

LSI 580-70
Fall 1999

CONTENTS:
1. Context for action research
2. Major concepts and principles of action research
3. Techniques of action research
4. The research notebook
5. Analyzing action research data
6. Ethical criteria for action researchers
7. The written action research report

3. TECHNIQUES OF ACTION RESEARCH

We will discuss four kinds of action research methology: 1) observational and narrative research methods, 2) non-observational, survey and self-report techniques, 3) discourse analysis and problem-solving methos, and 4) critical-reflective and evaluative research methods.

OBSERVATIONAL AND NARRATIVE RESEARCH METHODS

Styles of Observation

There are two styles of observation: participant obervation and non-participant observation. In participant observation, the researcher is a normal member of the group and joins in wholeheartedly with the activities, events, behaviors and culture of the group. In non-participant observation, the researcher is unobtrusive and does not engage in the roles and work of the group as a group member, but remains aloof and distanced from the action and does not feign membership of the group. In both styles, the identity of the researcher may or may not be known. In non-participant observation the researcher may be either visible to or hidden from the group being observed.

Recent developments in federal regulations governing research with human subjects may impact styles of observation. [I hope to have a guest lecturer discuss this topic later in the term. Human subjects research and Institutional Review Boards will be covered later in the term as well.]

Narrative Data

• Unstructured observational studies
  An example of unstructured observations is specimen records (stream-of-behavior chronicles or shadow studies). For specimen records, the researcher observes behavior in a natural setting over a brief period of time, such as a day, and records as much of the detail as possible. The result is a rich, detailed description of episodes.
• Structured observational techniques
  Three examples of structured observations are checklists, rating scales and interaction analysis protocols. Checklists permit the observer to 'check' to see if a pre-specified behavior is present. Rating scales use coded scale procedures to allow the observer to rate a particular behavior using graphic, numerical, or categorical rating scales. Interaction analysis protocols use pre-specified categories of verbal and non-verbal behavior, which are ticked during the cycle of observation. See your course workbook for examples of research instruments for structured observatnal techniques.

1. List points to be observed in their actual sequence of happening
2. Categorize all similar traits or attributes together
3. List all relevant and specific points of behavior
4. Provide space on checklist for tallying marks

NON-OBSERVATIONAL, SURVEY AND SELF-REPORT TECHNIQUES

ATTITUDE SCALES

Using attitude scales, the researcher is able to determine the strength of attitudes or opinions held by the respondents on a variety of attitude statements. By assigning a numerical code to each item on the scale, a composite can be determined representing the range of attitudes within the group studies.

The aim is to force resondents to study their attitudes and feeling about a number of issues.

QUESTIONNAIRES

INTERVIEWS

The interview has the advantage over the questionnaire of allowing the interviewer to probe areas of interest as they arise during the interview.

DISCOURSE ANALYSIS AND PROBLEM-SOLVING METHODS

CONTENT ANALYSIS

The goal of content analysis is to uncover hidden themes, concepts, and indicators of the message content. In its simplest form, some working categories are set up and the occurrences of category units are observed, counted, and noted. Content analysis could be performed on student logs or diaries of work on a research project.

EPISODE ANALYSIS

The aim of episode analysis is to establish coherent units and to examine relationships among and between these units, or 'episodes,' of classroom behavior. Episode analysis could be performed on a discussion of steps in the research or information-seeking process.

BRAINSTORMING

GROUP DISCUSSION

CRITICAL-REFLECTIVE AND EVALUATIVE RESEARCH METHODS

TRIANGULATION

Triangulation involves selecting a combination of methods of data collection. Typically it is a combination of observation and interview collected from three perspectives. One form of triangulation is to build relationships among the observations of the student, the teacher, and a critical friend (a neutral outsider invited to support data collection). Triangulation could also be among the observations of the student, the classroom teacher, and the library media specialist. Another form of triangulation is to build relationships among data collected using various techniques, such as stream-of-behavior chronicles, questionnaires, and interviews. Triangulation gives a more balanced picture of the situation and permits contradictions which are often hidden to become visible.

COLLEGIAL REVIEW

LECTURE FEEDBACK

STUDENT COURSE/TEACHER EVALUATION