As social life is increasingly mediated by technology, sites like Twitter, Facebook, and Reddit are becoming more useful sources of information on consumer behavior.
However, the data collected from these sites is unstructured – meaning it has not been structured in a pre-determined manner, the way survey data is.
Most qualitative research methods require the analysis of unstructured or semi-structured data: interview transcripts, field notes, or video journal entries.
However, these data forms tend to be fairly manageable in scope, as well as somewhat structured (ex. an interviewer usually follows a guide, so information will be structured according to that guide). Researchers are also exposed to data as they are collecting it, allowing them to conduct analysis as an iterative process that starts during the data collection process.
Digital data ultimately presents a different kind of challenge for qualitative researchers because they are not as deeply familiarized with their data during the collection process. The amount of digital data available can be fairly extensive and is nearly completely unstructured.
However, there are existing approaches to the analysis of qualitative data within the field of anthropology that are very helpful in managing digital data.
Although anthropologists are best known for their use of participant observation – immersing themselves in a cultural environment to participate and observe – collecting and analyzing “cultural artifacts” is another important facet of anthropological methods.
Cultural artifacts can range from household items (ex. a bowl) to tools (ex. a hammer) to written documents (ex. a newspaper article). To qualify as cultural artifacts, objects merely need to convey something about a specific culture.
We can also think of certain cultural artifacts as “texts” for analytical purposes. This type of terminology merely implies that an object – a written text like a paper, or an iconic text, like a drawing or painting – can be interpreted to elucidate meaning.
There are several different analytical traditions through which meaning can be inferred from texts:
Discourse analysis is an approach that treats language as a social interaction and emphasizes the importance of the social context in which discourse exists. It overlaps with semiotic approaches, which explore the meaning of signs and symbols in language and culture.
For instance, in a recent study, MDRG analyzed digital conversations to learn about how people buy a certain consumer packaged good. We noticed that certain consumers talk about organic and conventional food very differently. Contextualizing this discourse within a broader social context allowed us to recognize that discourse on organic and conventional food is less about the food itself than the relations between people. For some, organic is a way to convey a message of status centered on the idea of health or a way of mitigating discomfort with industrialized food production.
For others, particularly those of lower socio-economic status, organic is associated with high prices and upscale marketing that can feel alienating. Their notions of health are centered on other things, like time spent with family or balanced meals.
Content analysis involves the categorization and classification of text and can be used to quantify qualitative data. Qualitative coding software is increasingly playing a role in content analysis, allowing researchers to accelerate analytical processes. Nonetheless, computers are unable to replace the vital role that human creativity and critical thinking plays in analysis.
Ultimately, qualitative researchers must continue to cultivate traditional analytical skills in order to make sense of extensive and complex digital data.