Experiments are great but what I am starting to fall in love with is the field of secondary data analysis for research in education. As I began my Ph.D. work in education about two years ago, I had this image of a doctoral student with a bag around his shoulder and thick glasses, walking with a purpose in the dark empty hallways of an inner-city school going from class to class making observations, taking notes and pondering the reasons why things are the way they are. This image of a nomad searching for the truth was something of an exaggeration that my brain had developed through limited knowledge of what a doctoral education meant. I was set straight as soon as I set foot on the fourth floor of the Harrington building to go to my new cubicle at the office of Aggie STEM, you spent most of your time in front of a computer doing either data analysis of writing the results of that data analysis. And moreover, that data was not even collected by you.
I learned very quickly that research in education was divided into two paradigms, qualitative and quantitative. Qualitative research dealt with talking to people and observing them and taking notes and then reading those notes and then doing a lot of stuff with those notes and then coming to some conclusions. Quantitative research on the other hand deals with numbers. You find data represented in terms of numbers, you analyze the data, come up with some observations, develop some questions, use the data to answer those questions and then make conclusions with a very high accuracy defined as the p value. It was very clear that most people (at least those who I was going to work with) were interested in the quantitative side of the research. And now that I have immersed myself in this research it kind of makes sense.
The kind of data analysis where you actually did not conduct an experiment to produce the data or if the data was attained using a survey or if it is historical data collected as a result of an event such as an election or a state wide test or data produced by an entity regarding itself such as Texas Education Agency producing data about its schools is called secondary data. This data can be available publicly or it may be available by request. Either way there is a ton of it available and I feel, no I believe that every doctoral students’ first research project should be a to take a secondary data set and analyze it using a computer statistical software.
Analyzing a secondary data set allows one to enter the world of the others. The insights one can get about an entity or a group through secondary data analysis are always interesting since they are usually unexpected and often enlightening. For example, I along with a colleague recently analyzed a data set available online which concerns the K-12 teachers who are engaged in STEM (Science, Technology, Engineering and Mathematics) subjects. The data set is rich and is based on a lengthy survey. We chose to ask questions regarding STEM teacher retention based on school and classroom level factors and job satisfaction. When we were done analyzing the data set we didn’t just find the answers to the questions we were asking, but we were also able to come up with some new question. Some would be answered by analyzing the data set further but to answer the others we might have to search for a new data set or and here is the kicker, we may have to design an experiment or interview some teachers.
Aamir is a PhD student in the College of Education and Human Development