Ettema – 11/7

Reflections on Assignment 1 – Ettema will look for strong conclusions, rather than perfect questionnaires; even if McKinsey is doing your research, you should be involved in the action plan that results from it.

Case Study Research

What did we talk about at the beginning of the year? Everything is a variable, and we should be able to conceptualize variables. Change and Difference = Variation.

What else would make sense to do, besides Variable Analytic Research? In explain, predict, control, how else could we explain things? (Explanation = Show correlation between variables, show temporal order, and 3rd variable) When we explain in everyday life, we tell a story. Telling a story is the Great Alternative! Two ways to explain:

Quantitatively – show a relationship between measurements
Qualitatively – Tell a story

A story is about a change toward a difference; temporal order is built in. A description of a static situation is not a story, it doesn’t explain a change, or how/why change occurred.

How about prediction? Can stories predict? Does someone else’s story predict mine? What is it we get out of this method? We want knowledge claims that stand up…we’re presenting to “worst enemy” in a meeting…what would they say about case study research? “What is your story good for?”

Case studies are good for answering the “how” question. How something happened, the answer to the why question. When we study our expert, the question we should ask is how does he/she do their thing. (The “Why” is because it’s their job)

A real world example of not doing qualitative research…

The chief engineer at Ford during development of the Mustang tells this story: When Ford decided in the 70s to build a new kind of car, and prepared to produce the Taurus, they asked themselves, when was the last time we did this? It was the Mustang. They wanted to organize their efforts in a similar way for the Taurus, but they didn’t know, so they had to “reinvent the wheel.” A record of the “Hows” would have been helpful.

Quantitative data will tell us Why, but Qualitative tells us How.






Qualitative also provides meaning…less tangible, more symbolic explanation.


What is the method, or how do we tell a good story?

There isn’t as clear, rule-bound a method to qualitative research…more creativity, fewer rules. Jim puts qualitative and quantitative research on a continuum of subjectivity/objectivity. (e.g., Questionnaire writing is a subjective, creative process) We have to use own subjectivity to interpret responses, a human relationship must exist to conduct the research.

In case study research, more rigor and discipline is needed to make conclusions stand up…we have to make our own rules in a rigorous, disciplined way, since the method is looser. The greater element of subjectivity must be brought under control.

The conceptual vocabulary – A set of ideas or concepts, and a setting in which we study them. Jim’s CTW study…Sesame Street was popular and delivered on its educational aims. It tried to close the gap between lower and middle income kids. Two approaches: first was to teach basics and repeat. The other approach was to gauge what the school system demanded. (a “civility curriculum” existed to help lower income students relate to the college-educated teacher…”gaming the system”) Sesame Street was the great model…what else can be done? An idea for a show for older kids taught leadership, assertiveness…difference between boys and girls. Harold Lesser (book “Creation of SS”) argued that SS had a loose, organic collaboration between individuals without an underlying organizational theme…”temporary team.”

Ettema found a good answer to the how question, how SS was made? The group was designed to work in a particular way. More teachers sat on the creative committee than any other, thus educational values were prioritized. A researcher chaired the committee. The committee was not making decisions, and said to the producers “we trust you, you can make the decisions.” The committee realized it couldn’t legislate, never took votes. Jim couldn’t decide how it worked, needed an idea, or conceptual vocabulary, before designing his research, he found it in organizational power literature. He researched power, governance. In SS, power was accepted by producers, to the relief of the teachers, who wanted influence only. Jim realized he’d been attending the wrong meetings…

The key ideas were Strategic Contingencies of Organizational Power. Who has power in an organization? E.g., some VPs are more powerful than others, though their titles are the same. People who manage the strategic contingency are the ones with power; the difficult, important function creates the power. (AT&T before/after 1984…before, engineers were powerful when reliability was the only issue and no competition existed, after the divestiture, marketing/sales was more powerful) Non-substitutability is another element; if only one person can do something, he/she has power.

The producer had power because he could get a show in the can that kids would be willing to watch; writing educational plans was easy. Also, no one else could produce the show well; he had experience producing shows, i.e. he could listen to teachers, and in some respects duplicate their efforts. (The teachers would make small changes to scripts, suggestions) The teachers couldn’t duplicate his efforts.

Because the conceptual vocabulary is important, Jim suggests:

We select our expert:

1) a person with strategic value to us
2) a person whose job we want


The Schon reading will help! He hands us a conceptual vocabulary. (Most people hate him, but a few say, “it changed my life”) Ettema/Glaser Chapter 2 will give us what Schon is saying, demonstrates what we are trying to do in the project. (Chapters 3 and 6 are also worth reading) The other metaphor for conceptual vocabulary is that it’s a set of dissecting tools…the way we examine the interts of the subject animal.

We only need a paragraph on the expert’s life…Jim want the hows of the expertise the person demonstrates. We should enter the process with an open mind, but with a full tool kit/concepts and ideas.

James Caulstrom is an airline crash investigator…he says he has “every hypothesis” when he begins an investigation.

If we partner on this project, we should do two cases and do a comparative case study.

Our expert is not necessarily someone famous, but someone, “useful.”


We need a good way to get an answer to the How question…

How is Charlie Trotter’s wait staff trained? We articulate more specific questions from that general one…(Schon helps, e.g. on the spot experiments, how does the expert know what to look for?)

We have to make rigorous and disciplined the non-rule-bound data collection process, to get a credible story, one that will pass the test with boss, enemy in a meeting.

We’ll study:

1) Units of Analysis – Focal Units, the story we want to tell, in a series of episodes –could be individual, but it will be the role of the expert individual, group, organization, or industry for our project. (Tracy Kidder’s “The Soul of a New Machine” shows a group as unit) It could be as large as societies, the globe…to go the other direction, smaller than individuals, the episode. (episodes can be further broken down into a series of exchanges) Jim likes example of sales reps…what episodes of sales reps activity do we see? Practices connect episodes with a pattern. “Capturing and recording best practices.” No rules exist, like in accounting…“The enactment/practice of the sales role, as told in three episodes.”
2) Single vs. Multiple Cases – multiple cases provide generalizability, predictability, and repeatability, e.g. three best sales reps could be studied. What’s the right number of cases? Best Answer - We keep doing cases until we don’t see anything new, i.e., ”enough.” We get the range of possibilities. Second best answer – 3. Since we need “credible knowledge,” three cases provide credibility. Multiple cases might be repetitions, but they are better viewed as comparisons. With multiple cases, we can compare/contrast cases.
3) Choice of Particular Cases – Which cases to analyze? - One might select the “best” in their business, or best vs. mediocre for comparison…a Purposive Sample, we choose people not at random, but based on a clear purpose. “Justify our expert.”
4) Researcher’s Role – “Complete Observer” – e.g., watching lawyers in a courtroom…its opposite is the Complete Participant – “Observer Participant/Participant Observer” – part of activity observed, probably our approach.
5) Data Collection – three kinds of data 1) Observation 2) Interviews – in depth 3) Other (records, documents, materials that already exist) Find an expert that we can observe, and create a method of observation, such that we can say with confidence that we have seen him or her in action. “Protocol Analysis” – ask people to tell us what they are doing, as they are doing it…or talk before or after about what happens. For interview, the “tell me how it’s done”, we have two options, the interview funnel…ask big questions first, then funnel down to specifics. (always ask for examples) We may need to go the opposite way, start with specific questions to get to broader concepts. (Ettema/Glasser asked for the story of the award-winning article or book, interrupted at key points, when they said they didn’t know what to do next, sought guidance from editor…then, when they’d earned the right to ask, they drew out answers about “the truth.”)
6) Analyzing Data – Take all material, and see whether it converges, or diverges. (We might talk to the expert’s followers, and see if all agree, or not) Look for convergences, but don’t shy away from divergences.

What do we get out of the method at the end? What is the story worth?

- Problem of Genralizability is always there, but we get Carefully Analyzed Experience, looking for interpretation and meaning. We’ll always be able to say what we as the analyzers say, and show the interviews and observations that pointed us to our conclusions. We get more than data, we get Wisdom! (We have 2 weeks to get wise) Wise people don’t predict the future, they have a sense for situations, and they provide framework/basis upon which to act. “Reflective Practitioner…What makes one wise?”

- posted by Robert @ 2:52 PM