A research question is the fundamental question inherent in any research topic under investigation.
Qualitative research questions are typically open-ended, providing for a holistic view (ask how or what). Quantitative research questions are more focused, usually on only a few variables.
Research questions should not require a simple yes or no answer.
Research questions should not be stated in a manner that assumes an answer before data have been collected.
Research questions should be based in the body of literature related to the topic.
Research questions must be able to be answered by collecting available data.
Research questions must be ethical.
Research may be unethical if the study exposes participants to any physical, emotional, and psychological risks of any kind.
Research questions must be important and feasible to answer.
Research questions should truly reflect the topic being studied.
Hypotheses are tentative but intelligent, informed predictions about the findings of a study.
There are three types of hypotheses: null hypotheses, nondirectional research hypotheses, and directional research hypotheses.
The null hypothesis states that no effect, difference, or relationship will be found between variables.
The nondirectional research hypothesis states that an effect, a difference, or a relationship will be found but does not specify the direction of the effect, the difference, or the relationship.
The directional research hypothesis also states that an effect, a difference, or a relationship will be found and specifically indicates the direction of the effect, the difference, or the relationship.
A research design is the basic blueprint for conducting an action research study.
Qualitative research designs are less structured and more holistic in their approach to conducting a study than are quantitative designs.
A case study focuses on the detailed examination of a single setting, a single subject, or a particular event.
In observational studies, the researcher becomes an integral part of the setting of the study. The researcher may participate as an observer, an observer as participant, a participant as observer, or a full participant.
The constant comparative method is a qualitative research design for studies involving multiple data sources, where data analysis begins early in the study and is nearly completed by the end of data collection.
Quantitative research designs fall into four categories: descriptive designs, correlational designs, group comparisons, and single-subject designs.
Descriptive designs simply attempt to describe the current status of the phenomenon of interest. Descriptive designs include observational research and survey research.
In observational research, as a quantitative design, the focus is on a specific aspect of behavior, perhaps a single particular variable.
Survey research involves acquiring information from individuals representing one or more groups--perhaps about their opinions, attitudes, or characteristics--by specifically asking them questions and then tabulating their responses
Correlational designs investigate the extent to which a relationship exists between two or more variables.
Correlation coefficients report two aspects of the relationship between given variables: the direction of the relationship and the strength of the relationship. The strength of the relationship is indicated by the magnitude of the numerical value of the coefficient.
A positive correlation indicates that as the scores or values on one variable increase, the values on the other variable also increase, and vice versa.
A negative correlation means that as the values on one variable increase, the values on the other variable decrease.
The results of a correlational study cannot be used to explain causation.
Group comparison designs examine cause-and-effect relationships and involve a manipulated independent variable and a dependent variable measured across all groups.
Group comparison designs include causal-comparative designs (which explore the cause of an effect after the fact), preexperimental designs (which typically involve one group simply being “compared” with itself), and quasi-experimental designs (which involve two groups being compared with each other on a common dependent variable).
Mixed-methods research designs involve the collection of both qualitative and quantitative data.
Many action research studies tend to “align” better with mixed-methods research designs than qualitative or quantitative methods alone.
In an explanatory mixed-methods design, the educator-researcher first collects quantitative data and then gathers additional qualitative data in order to help support, explain, or elaborate on the quantitative results; the focus is on the quantitative data.
In an exploratory mixed-methods design, qualitative data are collected first, followed by quantitative data, in order to further explain any relationships discovered in the qualitative data; emphasis is placed on the qualitative data.
In a triangulation mixed-methods design, both quantitative and qualitative data are collected at about the same time and are given equal emphasis.
Ethical treatment of students and colleagues--as well as their respective data--must be a key component of designing your action research study.
Most educational resource must go through a review process. These are typically conducted by review boards such as Human Subjects Review Boards (HSRBs) or Institutional Review Boards (IRBs).
There are several training modules for preparing to address ethical issues in educational research.
Issues to consider regarding the ethics of a research study include participant permission or consent, confidentiality, the principle of beneficence, the principle or honesty, and the principle of importance.
Getting organized for your research can include: integrating action research within the context of what you typically do in the classroom, developing a research schedule, and maintaining flexibility in the research schedule and topic area.