Your measurements, or dependent variables, provide the data you will analyze. There are two main types of data.

Quantitative data represent the world with numbers that can be statistically analyzed. They are necessary for relational or causal research questions because these types of questions require inferential statistics that generalize the results from your study to a larger population (more on this later). For example, quantitative data could tell us the average number of forum posts per student or the number of times students watched a video. Quantitative measures are appropriate when you want to confirm a hypothesis (e.g., that students in one group outperform others), but they are close-ended, which does not allow for exploration.

Qualitative data, in contrast, are open-ended and used mainly for descriptive research questions. For example, qualitative data could tell us what a student posts on a forum or notes a student took while watching a video. Qualitative measures are appropriate when you want to explore a phenomenon (e.g., how students use forums), especially if you want that exploration to be driven bottom-up by the learners rather than top-down by the researchers. However, qualitative data can be highly detailed and time-consuming to analyze. As a result, many researchers will analyze qualitative data quantitatively to summarize the data for future readers. Qualitative data can be quantified using coding schemes that turn descriptions into numbers. For example, for coding a forum in a physics class, you could use a coding scheme that counted the number of times students mentioned each of Newton’s Laws of Motion. Quantifying data allows qualitative data to be used in statistical analyses and for relational and causal questions.

A growing group of researchers leverages the strengths of both quantitative and qualitative data by employing a mixed methods design. Of course, this approach is not appropriate for all research studies, but developing skills in collecting and analyzing both types of data, or finding collaborators, affords a wide range of options. The most common textbook on mixed methods designs is from 2007 by Creswell and Clark. They describe four typical designs for mixing quantitative and qualitative data.

Mixed Design	Order of Quant/Qual	Mixed Goal	Interpretation
Triangulation Design	Qual results used to develop a theory or instrument to test quantitatively	Compare results from each type to interrelate and validate findings	Equal emphasis on both types of data
Explanatory Design	Quant followed by Qual	Quant results require further clarification or participant selection	Emphasis on quant data will qual supporting explanations and elaborations
Exploratory Design	Qual followed by Quant	Qual results used to develop theory or instrument to test quantitatively	Emphasis on qual data with quant demonstrating generalizability
Embedded Design	Qual embedded in Quant (or vice versa)	Qual data enhance the quant results (or vice versa)	Emphasis on quant data (or vice versa)

To view more posts about research design, see a list of topics on the Research Design: Series Introduction.