By Isha Dandavate, Sophie Barness, Seema Hari

Big data analysis has become a widely used tool– Netflix used it to inform the creation of “House of Cards,” Nate Silver used it to predict the results of the presidential elections, and Grameen Foundation uses it to better tailor financial services for poor households. However, big data comes with its own set of problems. This New York Times op-ed piece, by David Brooks, discusses situations in which data doesn’t work; and coincidentally these are situations in which the strengths of qualitative research could be leveraged.

Brooks points out a basic problem with computers (it’s one we should all be familiar with): the sociotechnical gap. In his words, “Computer-driven data analysis… excels at measuring the quantity of social interactions but not the quality.” So where data might tell you that in the last week, you have had five conversations with a particular classmate but 15 conversations with another, it wouldn’t tell you that actually, the first classmate is your best friend who happens to have a sick child at home, whereas the second classmate is your frenemy with whom you are working on a group project. If a company used conversation data to automatically set privacy settings, we would have a problem on our hands.

In this example, quantiative analysis would test the correlation between number of conversations and closeness of the friendship. However, qualitative methods take a different approach. Becker wrote in “Epistemology of Qualitative Research,” that “the point is not to prove, beyond a doubt, the existence of particular relationships so much as to describe a system of relationships.” (352) So, where data can’t always accurately explain social dynamics, qualitative data might be able to better capture the nuances of social interaction– largely because we don’t rely on a computer program to process the information, and instead rely on conversations or observations of a researcher (who possesses the “machine in your skull” that Brooks refers to).

Brooks writes that big data has trouble with big problems. When addressing issues where experimental control situations are applicable data can tell us a lot, but data has a hard time trying to figure out big problems in society where no alternate “controlled” society can be studied. Qualitative methods can often fill in these gaps through observation of everyday situations.

Qualitative analysis considers the breadth and full description in order to create an accurate picture. Becker wrote that the best type of research is one “based on careful, close-up observation of a wide variety of matters that bear on the question under investigation”. So where quantitative analysis tries to take the chaos of data and make it more simple, Qualitative analysis approaches bulk and chaos by creating frameworks that would best encapsulate the scenario at hand.

According to Brooks, big data misses the dimension of context which is a key aspect of the decision making process. Qualitative data fills this gap. Becker discusses how ethnographers “observe people when all the constraints of the ordinary social situations are operative,” which adds context information to the data that can be collected about people. Researchers can recognize many of the environmental variables and emotions at play as and when the human decisions happen, adding the critical information that quantitative data misses.

Furthermore, humans are much better at telling their stories which weave together multiple contexts and data analysis excludes this narrative style. The results of qualitative research are presented in a very descriptive way through personas, storyboards, opportunity maps and scenarios which replicate the way the humans tell their stories. Personas add context to the demographic information of a target user and could help designers make decisions by putting themselves in the persona’s shoes. It plays on the cognitive ability of the designers being able to recognize and act like the personas while making design decisions.

In his assessment, Brooks addresses the nuances of the human experience that big data can’t possibly capture, specifically, social values, complexity, and context. Qualitative research, by the virtue of its ethnographic methods, is able to capture these; as we read, social mapping is the end goal, complexity is unavoidable and thus a tool by which ethnographers refine their findings, and qualitative findings are only valuable when presented in context. So perhaps the takeaway here is that, these methods are complementary– and if we can employ both qualitative and quantitative findings to any given problem, it may enhance our ability to accurately describe human experiences in a way that could impact business, politics, and development.