What do you mean by Semi-structured Analysis?

We know of primarily two categories of Analysis, the structured and the unstructured analysis. The semi-structured analysis lie somewhere between the both. While a structured analysis is based on data collected from closed-ended questions, the unstructured analysis is more from observed data and no pre-decided questions. Questions usually come up spontaneously during the observation.  When we talk of semi-structured analysis, it talks of data with open-ended questions where the respondent gets the freedom to answer the questions in a way he thinks most appropriate. It adds an element of subjectivity to the responses. The use of semi-structured analysis is seen more in explorative or qualitative research where the existing data available for study is limited or does not exist. The motive of the researcher is to surface out the underlying reasons or motivations that are responsible for some kind of behavior. It is imperative the researcher has very worthy communication skills because the respondents have to feel comfortable and safe in the presence of the investigator else they would not be able to express their opinion and choices with complete honesty. Usually, it works well in an environment where the researcher and the respondent, both belong to a familiar background to each other so understand each other’s comfort level. With this limitation of semi-structured research, when a researcher has to expand his horizons and move to a diverse culture for the purpose of research, the help, and involvement of interface researchers should be sought. This is because the respondents don’t feel constrained in the presence of an unfamiliar researcher, and their responses don’t become biased. This may ultimately affect the output of the research.

The suitability of semi-structured analysis lies in a situation where the researcher knows something about the difficulty but still there is a lot more to be explored. In a situation, where the researcher has no knowledge about the challenge, he must first begin with an unstructured analysis and then after generating some familiarity, frame relevant questions for repeat research which would be the semi-structured research at this stage. This research sets frontier for respondents, and they have to restrain themselves within that frontier. The output, in the form of collected data, would have both associable as well as unfamiliar responses. The scholar would have to do filtering at this stage so that what is unknown can be further explored.