The program you see today is the result of rock-solid research by Edward Brent and Pawel Slusarz a few years back. Their seminal paper “Feeling the Beat: Intelligent Coding Advice from Metaknowledge in Qualitative Research” outlines their strategy for increasing the efficiency and quality of coding textual data by tapping into the great store of knowledge gathered by the coding process itself.
While computers offer a vast improvement over coding data by hand, the coding process isn’t eagerly anticipated by most researchers because it still requires vast manual effort:
In general, the work required for the coder for the second segment to be coded, or even the thousandth segment to be coded, is as great as it was for the first segment….The level of knowledge and the attention to detail required of the coder is as great or greater near the end of the project as it was at the beginning. As a result, coding is widely recognized to be a tedious, detailed, repetitive, even mind-numbing task.
Brent and Slusarz realized the metaknowledge gathered throughout the coding process could reduce the burden to researchers:
Between the time when the first segment is coded and the time when the last code of the project is entered, the researcher has amassed a vast database of coded data, typically including hundreds or even thousands of segments of text (or other materials)….The patterns of data that result have a clear structure, a “rhythm” if you will, that provides the underlying “beat” within which the “melody” of codes is experienced. Together, these diverse elements of knowledge about the data (the metaknowledge) provide a surprisingly informative set of cues as to what codes might be expected for any segment.
The authors go on to show how intelligent computational strategies—case-based reasoning, natural-language generation, semantic networks, and production rules—can take advantage of the knowledge implicit in coded information in qualitative databases to help code additional data.
If you’re interested in learning more about how Qualrus utilizes intelligent computational strategies to make coding easier, check out the full article: Feeling the Beat: Intelligent Coding Advice from Metaknowledge in Qualitative Research.
(Originally published in Social Science Computer Review by Sage Publications)
]]>As a researcher or analyst you probably know a thing or two about the value of qualitative data, but here is a refresher. And hey, maybe you’ll learn something new too.
One of the key differences between open-ended data and closed ended data is the responded has the freedom to respond in any fashion desired. Many times this leads the researcher to discover issues not explicitly anticipated before hand.
Qualitative data and more specifically with open-ended surveys a researcher can address more issues, including less common ones, without having to increase the number of questions asked.
Often times with qualitative data a researcher discovers fine grained detailed information that brings up issues in a new perspective that may not happen with quantitative measures.
Sometimes overlooked, how a respondent feels is often quite important to the type of response a researcher will receive. Open ended surveys allow respondents to express concerns in their own voice at lets them know someone is really listening.
Along the same lines, qualitative data demonstrates to the respondent a commitment to pay real attention to responders comments.
Additionally and traditionally, qualitative data has worked well for discovering popular and new issues. These issues can then be incorporated later into future quantitative measures.
In just these few points you can see just how powerful qualitative data really is. In addition to a few quantitative metrics on the issues you’re looking at, you can get a true glimpse at your data and discover all the details that you’re looking for and even some you aren’t.
]]>You may have heard the world is made up of atoms and molecules, but it’s really made up of stories. When you sit with an individual that’s been here, you can give quantitative data a qualitative overlay.
]]>