Types of Education Research

There are two basic types of education research: descriptive research and experimental research. Each type answers different research questions and uses different research designs to collect data.

DESCRIPTIVE RESEARCH QUESTIONS AND DESIGNS

Descriptive research is used to answer descriptive research questions: What is happening? How is something happening? Why is something happening?

Examples:

• What is the average number of staff development hours per year for teachers in the United States?
  
• What is the association between student-teacher ratios and student achievement in the state’s elementary schools?
  
• How does instruction differ among teachers in the district who receive different amounts of staff development?
  
• Why do teacher qualifications influence instruction?

• Simple descriptive
• Comparative descriptive
• Correlational

A simple descriptive research design is used when data are collected to describe persons, organizations, settings, or phenomena. For example, a researcher administers a survey to a random sample of teachers in the state in order to describe the characteristics of the state’s population of teachers.

With a comparative descriptive design, the researcher describes two or more groups of participants. For example, a researcher administers a questionnaire to three groups of teachers about their classroom practices. The researcher chooses the three schools because the schools vary in terms of the amount of professional development that they provide to teachers.

A correlational research design is used to describe the statistical association between two or more variables. For example, a researcher measures the student-teacher ratio in each classroom in a school district and measures the average student achievement on the state assessment in each of these same classrooms. Next the researcher uses statistical techniques to measure whether the student-teacher ratio and student achievement in the school district are connected numerically; for example, when the student-teacher ratio changes in value, so does student achievement. The researcher can then use the student-teacher ratio to predict student achievement, a technique called regression analysis. When there is more than one predictor variable, the technique of multiple regression analysis produces a multiple correlation that is used for prediction.

EXPERIMENTAL RESEARCH QUESTIONS AND DESIGNS

Experimental research is used to answer causal research questions: Does something cause an effect? For example, does a low student-teacher ratio cause higher student achievement?

Experimental research designs include the following:

• True experimental (randomized trials)
• Quasi-experimental

In experimental research, the researcher manipulates or varies an independent variable and measures its effects on one or more dependent variables. In a true experimental design, the researcher randomly assigns the participants who are being studied (also called the subjects) to two or more comparison groups. Sometimes the comparison groups are referred to as treatment and control groups. Participants in the treatment group receive some type of treatment, such as a special reading program. Participants in the control group do not receive the treatment.

For example, at the beginning of a school year, a researcher randomly assigns all classes in a school district to have either a low student-teacher ratio (small class, the treatment group) or a normal student-teacher ratio (large class, the control group). At the end of the school year, the researcher measures each student’s achievement using the state assessment and compares the average achievement of students in the two sizes of classes. In this example, class size is the independent variable because class size is being varied or manipulated. Student achievement is the dependent variable because student achievement is being measured. (Note: Researchers conducted a similar experiment in the state of Tennessee starting in 1985. The study is known as Project STAR.)

In a quasi-experimental design, the researcher does not randomly assign participants to comparison groups, usually because random assignment is not feasible. To improve a quasi-experimental design, the researcher can match the comparison groups on characteristics that relate to the dependent variable. For example, a researcher selects from a school district 10 classes to have low student-teacher ratios and 10 classes to maintain their current high student-teacher ratios. The researcher selects the high-ratio classes based on their similarity to the low-ratio classes in terms of student socioeconomic status, a variable that is related to student achievement.

For a more in-depth discussion of experimental research, refer to the publication Identifying and Implementing Educational Practices Supported by Rigorous Evidence: A User Friendly Guide, recently released by the U.S. Department of Education’s Institute of Education Sciences. The publication can be viewed on the Web at http://www.ed.gov/rschstat/research/pubs/rigorousevid/index.html or downloaded at http://www.ed.gov/rschstat/research/pubs/rigorousevid/rigorousevid.pdf.

Data and Data-collection Strategies

TYPES OF DATA: QUANTITATIVE AND QUALITATIVE

In quantitative research, the data are numbers and measurements; in qualitative research, the data are narrative descriptions and observations. Other differences are that qualitative research occurs in more natural and less controlled research settings than does quantitative research, and qualitative research often uses special methods to collect data, such as case study and ethnography. These methods reflect the philosophy of qualitative research, which emphasizes in-depth descriptions of persons, behaviors and contexts.

With regard to research designs, correlational, experimental and quasi-experimental designs usually collect quantitative data. Simple descriptive and comparative descriptive designs collect either type of data. When both quantitative and qualitative data are collected in the same study, the approach is called mixed methods.

DATA-COLLECTION STRATEGIES: LONGITUDINAL AND CROSS-SECTIONAL

Longitudinal and cross-sectional are data-collection strategies that can be used with either descriptive or experimental research designs.

Example of a descriptive longitudinal research study:
A researcher studies the relationship between the average class size that each student experienced in grade 2 and each student’s achievement in grades 2, 4, and 6. The purpose is to determine whether the relationship between class size and achievement remains the same or changes over four school years. In longitudinal studies, the emphasis is on individual change over time.

Example of an experimental cross-sectional research study:
A researcher randomly assigns 2nd graders, 4th graders, and 6th graders to classes that are either small or large in size. The purpose is to determine at the end of the school year whether the difference in student achievement between small and large classes varies depending on the grade levels of the students. In cross-sectional studies, the emphasis is on differences between groups at one point in time.

Action Research

Action research is a special type of education research that involves the following features:

• Collection of data about a current education practice or program and the resulting outcomes
• Reflection on the information acquired
• Development and implementation of an improvement plan (the action)
• Collection of data on the practice or program after changes have been made
• Development of conclusions about the results of the improvement plan

Action research can be conducted by individuals such as teachers or by groups of individuals, such as school staffs. The latter is called “collaborative action research.”

Example of individual action research:

A teacher wants to know if group activities will improve the performance of her students in math. She measures the performance of her students on math problems after using whole-group instruction for three weeks. She then supplements her instruction with small-group learning activities for three weeks and again measures student performance. She finds that student achievement increased with the use of group activities compared to whole-group instruction. On the basis of these action research results, she changes her approach to teaching math.

Example of collaborative action research:

The teachers and principal of an elementary school want to improve their students' writing skills. Together they examine student writing samples and identify the specific areas that need improvement. They then purchase a new writing program that teachers, subsequent to training, will implement for eight weeks. During the eight weeks, the teachers and principals meet weekly as a group to discuss progress and problems. At the end of the eight weeks, they compare student writing samples to those obtained prior to implementation of the new program. Because there are improvements in student writing, the school decides to adopt the program as part of their regular curriculum. The teachers and principal agree to continue monitoring the implementation and results of the new writing program.

For practitioners, action research can have several benefits. These include reflection on education practice, identification of strategies for improvement and acquisition of research skills. Collaborative action research has the additional benefit of engaging teachers and principals in joint work to improve education outcomes.

There are several limitations to action research, however. Theoretically, action research can be either descriptive or experimental. Most action research studies use descriptive research designs but attempt to draw conclusions about the effects of an action on some outcome. Action research studies rarely employ experimental methods, such as the use of a control group or the matching or random assignment that give experimental studies their power. Conclusions about cause and effect are reliable, however, only when they are based on solid experimental research designs. Another limitation is that most action research is restricted to one classroom or school, which means that the results cannot be generalized to other classrooms or schools. Thus, action research studies often lack both internal validity and external validity , and generally are not useful for making policy decisions.

References and Resources

Calhoun, E. F. (1994). How to use action research in the self-renewing school. Alexandria, VA: Association for Supervision and Curriculum Development.

Creswell, J. W. (2002). Research design: Qualitative, quantitative, and mixed method approaches. Thousand Oaks: Sage Publications.

Institute of Education Sciences. Identifying and implementing educational practices supported by rigorous evidence: A user friendly guide. Washington, DC: U.S. Department of Education.

McLean, J. E. (1995). Improving education through action research: A guide for administrators and teachers. Thousand Oaks, CA: Corwin Press, Inc.

McMillan, J. H. (2000). Educational research: Fundamentals for the consumer (3rd ed.). New York: Addison Wesley Longman.

National Research Council. (2002). Scientific research in education. Committee on Scientific Principles for Education Research. Shavelson, R. J., and Towne, L., Editors. Center for Education. Division of Behavioral and Social Sciences and Education. Washington, DC: National Academy Press.