Case Studies involve gaining detailed, intensive knowledge about a single instance or a set of related instances. These instances, or cases, may be of individuals, organizations, communities, events, or processes.
Case Studies is Day 11 of 100 Days of UX, an exploratory effort to survey the 100 methods of design outlined in Martin and Hannington's Universal Methods of Design. For 100 consecutive days, I learn one new method a day and write about it.
Performing a case study requires the researcher to discover a problem, devise hypotheses, validate (or invalidate) those hypotheses via interviews, observations, etc, revise the hypotheses, and tell a story. In this way, case studies bear some resemblance to the design process.
Case studies are useful in exploratory research for understanding existing phenomena as well as for studying the effects of change, new programs, or innovations.
A proper case study ought to be comprehensive and cover not only the instance under consideration, but interrelationships betweeen the instance and other things that may have a role to play in that case. Also, case studies should not only cover representative instances, but also those that significantly deviate from the norm.
The deliverable at the end of the case study process is a story. This story should be like a Justice Antonin Scalia opinion: literate, accurate, and useful for research, but also alliterative, enjoyable, awe-inspiring, and filled with vivid details that make it a joy to read.