Data

Forms of Data in Educational Research

Data can take on many different forms and have varying amounts of information attached to them. They can be qualitative or quantitative. Qualitative data is generally the term used for information that is not in the form of numbers or quantities. It might be written or spoken words or descriptions of observed behavior, or in a variety of other forms that cannot be easily summarized as amounts of something.

Quantitative data are in the form of numbers. These numbers, usually in the form of scores, can represent different levels of measurement and can be nominal, ordinal, interval, or ratio data. Moving from nominal to ratio levels of measurement will naturally carry more information with the data. Nominal data simply represent categories, whereas ratio data are quantitative with a meaningful zero.

Many variables in education are nominal, ordinal, or interval of nature. Understanding the level of measurement the data take is an important step because many inferential modeling procedures such as t tests or regression assume that the dependent variable is at least an interval scale of measurement.

One specific example commonly used in education is the collection of survey data through a rating scale (e.g., strongly disagree to strongly agree). It is common for educational researchers to represent these categories as numeric values; for example, if there are five categories, these are often represented as numeric integers from 1 (strongly disagree) to 5 (strongly agree). Sometimes responses to survey questions are summed to create a scale measuring some phenomenon; however, in other instances, the responses for a single item are used. Some caution should be exercised in assuming these types of data for a single survey question are interval in nature, as the gaps are likely not consistent across the entire range of the scale.

Using Data in Educational Research

Data in education can be collected from experiments, observational studies, or even computer [Page 458]simulations. These data are often used in descriptive or inferential analyses to help aid in making decisions regarding the effectiveness of educational programs. As computers have become more powerful and more portable, data can now be collected daily, hourly, or even every second. This has led to a significant increase in the amount of data available to educational researchers. In addition, more complex models are being used to explore this information that may make more assumptions regarding the data. Therefore, particular attention and time need to be made to thoroughly understand the type of data being used and whether assumptions are being met for the data analysis.

...