During data collection in research, interview is one of the tool adopted by the researcher; but has different forms such as structured, semi-structured and unstructured. How can we distinguish between them.
Different texts describe these structures slightly differently. Structured interviews are typically described as not deviating from the interview protocol, being delivered the same way to each participant being interviewed, etc. More or less of an in-person questionnaire. A semi-structured interview has an interview protocol, but it is also responsive to the answers participants give. In other words, it is flexible, allowing for unplanned follow-ups or additional questions that are raised in the context of a specific interview. An unstructured interview (again, generally speaking) is more of a free-flowing conversation, with a plan for basic topics but not necessarily a specific protocol. Resources for interviews vary widely by discipline, but one I have found useful is InterViews: Learning the Craft of Qualitative Research Interviewing by Kvale and Brinkman. I could also recomment Qualitative Interviewing: The Art of Hearing Data by Rubin & Rubin, Interviewing as Qualitative Research: A Guide for Researchers in Education and the Social Sciences by Seidman, and Research Interviewing: Context and Narrative by Mischler. I hope that helps get you started on sorting out those terms. Cheers!
I agree that Jessica's answer is very good; I'll just add that, in my opinion, while respectively, structured, semi-structured, and unstructured interviews put increasingly less constraint on respondents, at the same time, they put increasingly more responsibility on the researcher to collect valid data.
Technically, in a structured interview, an enumerator use a structured questions, all are of objective type. In semi structure, enumerator use a questionnaire having MCQ and open ended questions, in unstructured enumerator use fully open ended questions and use his /her personal observations also.
I think we should base on the method to distinguish structure, semi structure and unstructure questionnaire. If you apply the survey, it should be structure questionnaire, and if you conduct interview, they maybe semi or unstructure
One add to Jessica's answer is the number of interviewers. If there is a team of interviewers structured interviews may be preferred. Structured interviews helps eliminate the urge for interviewers to follow a "thread" that may be different than others' threads or on a tangent to the project's research questions. If you need to train additional interviewers keep in mind the different purposes and level of project goal understanding needed for the different types of interview protocols.
In addition to the "how to distinguish them" question, we can also ask and investigate "what is the difference between them in the efficiency of data generation". See, for example, the article:
M. Randle, H. Mackay., D.Dudley (2014) , A comparison of group-based research methods. „Market&Social Research” 2014, p. 25