Understanding Research Findings, Part 2: Research Designs. This work is being brought to you by researchers and educators at REL Central at Marzano Research. This presentation explains three basic types of research designs: Descriptive Research, Correlational Research, and Experimental Research. This is designed to assist educators in increasing their research literacy as they read existing research studies or consider a type of research that might help to shape decisions about educational issues in their community, state, or region. Let’s first look at descriptive research. Descriptive studies tend to look at existing conditions or trends in aspects of education. Examples of research questions addressed by descriptive research might be: a) What are teacher attitudes toward the use of a new mandated curriculum? b) What are current patterns of technology use in classrooms? c) How has the English Language Learner population changed over time? A limitation of descriptive research is that there is no information gathered about cause and effect relationships, so it may shed no light on why something is happening a certain way. Descriptive research is helpful to summarize data in order to describe what is happening. Results from descriptive studies can be used to inform future research by identifying variables that need additional testing or developing hypothesis about why things are happening. For example, if teachers report that they are dissatisfied with the new mandated curriculum, you might want to explore whether the satisfaction level is different for teachers with different years of experience, or if the professional development provided with the curriculum is effective. Correlational research describes the relationship between different variables. It does not say that one variable causes something to happen to another variable, but rather points to an association between two or more variables. Examples of research questions that might be addressed by correlational research include: a) How is parent involvement associated with students’ academic performance? b) Is the assignment of more homework associated with students’ academic engagement? c) Is taking classes in foreign languages correlated with student test scores in mathematics? Correlations can either be positive or negative. A positive correlation means that as one variable increases, so does the other one. For one of the example questions, is the assignment of more homework associated with students’ academic engagement?, a positive correlation would occur if researchers found that increased homework was linked with higher levels of academic engagement. Conversely, if students’ academic engagement decreased with more assigned homework, that would be a negative correlation. When one variable increases or decreases in value and a second variable does not change in relation, then there is no correlation between the two variables. Correlational research provides clear information about the association between two or more variables. The findings can establish a valuable starting point to researchers beginning to explore a subject. One limitation of correlational research is that it does not show what caused the relationship. The researcher may not have measured all the variables that are causing change; an unknown variable may influence variables that are being examined. If we consider an example in which researchers found a positive correlation between parent involvement and students’ academic performance, there are many possible explanations. It could be the case that parents who are more involved also have higher incomes and therefore their children may have access to more resources which could influence achievement. Experimental research is research that involves comparing two or more groups in order to draw conclusions about the causal effect of variables. Typically, one group receives a treatment while a second group, the control group, does not. The two groups are then compared at a later point to see how they differ on one or more outcome measures. Some questions answered by experimental research might be: a) Is this math strategy effective with fourth graders? b) Does access to mentoring support affect teacher retention? c) Does the peer leadership program impact student engagement and graduation rates? Experimental research can show whether an intervention is effective by comparing performance of a group who receives an intervention to a group who does not. In order to determine if an intervention is effective, the two groups being compared need to be as identical as possible. This is what allows causal conclusions to be drawn; the groups only differ on the independent variable, or the intervention being tested. Theoretically, any differences between the two groups is a result of the manipulation of the independent variable. In order to ensure that the groups are similar, experimental research often uses random assignment. This process is analogous to coin-flipping, where research subjects, such as individuals or classes, have an equal chance of being assigned to receive the intervention. By assigning groups at random the groups are likely to have similar characteristics. One limitation is that random assignment is not always possible, such as when students are already in classes and can’t be reassigned. In this case researchers can conduct a quasi-experimental study. A quasi- experimental study still features treatment and control groups, however since the groups were not randomly assigned researchers must control for variables that could be different between the two groups. One way to help control for this is through the use of baseline data to test for the similarity of the groups before the intervention was implemented. To summarize the three research designs: Descriptive research is used to understand a phenomena and to develop possible hypotheses about that phenomena. Correlational research is used to understand how variables are associated with each other, and it’s also useful to develop and explore hypotheses. Experimental research involves comparing two or more groups and allows researchers to draw cause and effect conclusions. For more information contact us at [email protected] or find information on these websites.