8 Research Design Case Study
Week 8:
TL;DR: , do your homework
readings (links) & lectures ～ assignments due ～ live session agenda
TL;DR: , do your homework
readings (links) & lectures ～ assignments due ～ live session agenda
Figure 26: Week 8 “reading” time estimated at 250 words per minute
Readings
Lectures
- ↓ 8.1 Introduction to Research Design Case Study Intro
- ↓ 8.3 Case Study 1: Using Data to Disrupt a Media Model
- ↓ 8.4 Case Study 2: Improving Relationships in Health Care
- ↓ 8.5 Case Study 3: Gambling on Data Insights
- ↓ 8.6 Case Study 4: Fitting the Numbers
- ↓ 8.7 Case Study 5: Insuring through Data
- ↓ 8.8 Video
8.1 Introduction to Research Design Case Study Intro ↑
| Players | Slug | Excerpt |
|---|---|---|
| π φ | beginning | We’ll start with an overview of the organization and its domain, then drill into the decision it faced and the questions it tried to answer with data. |
| θ ω | middle | We’ll learn more about that data and the efforts it took to structure the data to make it useful to answer the decision they faced. |
| δ τ | end | Think about also how you would convey findings to an audience, whether that’s maybe members of the people on your team, the organization itself, or the general public. |
8.3 Case Study 1: Using Data to Disrupt a Media Model ↑
| Players | Slug | Excerpt |
|---|---|---|
| α ι | Nflx problem | understand the best way to tailor a show to viewers’ preferences |
| ρ ω | Nflx RQ | What are our subscribers watching on Netflix today? What will our subscribers watch on Netflix in the future? |
| κ ρ | Nflx sub-RQs | How often has a particular movie or TV show been watched and by which users? How many repeat views have occurred for a particular movie or TV show and by which users? How many unique viewers have watched the most popular shows and movies? |
| ε γ | Nflx sub-sub-RQs | Is there a particular director whose work garners a lot of interest? Is there an actor who seems to drive more views? Are there shows from non-US markets that could be modified in a way that would generate interest within the market? |
| δ ω | Nflx results | there was a point of intersection between three major categories—a healthy share of customers had streamed David Fincher movies, Kevin Spacey received a lot of views, and the 1990 BBC miniseries House of Cards was a hit in the UK. |
| ψ η | Netflix inference | launching an American, David Fincher–helmed House of Cards starring Kevin Spacey would be a good bet |
| ρ υ | Nflx communication | Kevin Spacey fans saw trailers that featured him, women who watched Thelma and Louise saw trailers that featured the women cast members of the show, and serious movie buffs were treated to a David Fincher–themed trailer. |
| ο ρ | Nflx verification | There was also no in-depth analysis of whether the success of House of Cards could be attributed, to a meaningful extent, to the data model. |
8.4 Case Study 2: Improving Relationships in Health Care ↑
| Players | Slug | Excerpt |
|---|---|---|
| π δ | Gngr problem | proactively alerting the medical care provider before the patient himself or herself realized something was wrong |
| υ λ | Gngr ethics | In addition to tracking and analyzing the data, this approach also brought with it a slew of privacy issues that needed to be tackled. |
| ι θ | Gngr RQs | What sort of data is needed to help understand behavioral patterns that indicate whether patient behavior is outside of the ordinary? How can this data be collected unobtrusively? What is the correlation between the data collected and a patient’s behaviors or feelings? |
| λ ζ | Gngr tactic | a combination of patient-reported data as well as passive data from the phone’s sensors could help them predict a change in behavior |
| β α | Gngr communication | generate an alert to health care providers |
| ζ β | Gngr instrumentation | Creating an actual system required a combination of self-reported data from patient surveys and passive data from the phone’s sensors. |
| ξ τ | Gngr method? | generate behavior patterns whereby they could predict if a patient’s behavior or mood had changed. |
| α γ | Gngr client | clinical researchers were the intended audience for the Ginger.io data |
| ψ ο | Gngr unintended consequence | feedback indicated that these data were far more useful for health care providers who wanted to tailor their patient treatments |
| ρ ω | Gngr accomplishment | The tool now helps doctors track their patients’ behaviors and moods and thereby moves the conversation between health care provider and patient beyond that of a routine check-in. |
| α ι | Gngr extension | The company also plans to incorporate other relationships into this model, like the one between a patient and his or her supporter. |
8.5 Case Study 3: Gambling on Data Insights ↑
| Players | Slug | Excerpt |
|---|---|---|
| η π | NA | NA |
8.6 Case Study 4: Fitting the Numbers ↑
| Players | Slug | Excerpt |
|---|---|---|
| λ δ | NA | NA |
8.7 Case Study 5: Insuring through Data ↑
| Players | Slug | Excerpt |
|---|---|---|
| δ β | NA | NA |
Bibliography
Gitlin, Todd. 1985a. “Chapter 1 the Problem of Knowing.” In Inside Prime Time, 19–30. Berkeley: University of California Press. https://smile.amazon.com/Inside-Prime-Time-Todd-Gitlin/dp/0394737873/.
Gitlin, Todd. 1985b. “Chapter 2 Predicting the Unpredictable.” In Inside Prime Time, 31–46. Berkeley: University of California Press. https://smile.amazon.com/Inside-Prime-Time-Todd-Gitlin/dp/0394737873/.
Gitlin, Todd. 1985c. “Chapter 3 by the Numbers.” In Inside Prime Time, 47–55. Berkeley: University of California Press. https://smile.amazon.com/Inside-Prime-Time-Todd-Gitlin/dp/0394737873/.
Gitlin, Todd. 1985d. “Chapter 4 Making Schedules.” In Inside Prime Time, 56–62. Berkeley: University of California Press. https://smile.amazon.com/Inside-Prime-Time-Todd-Gitlin/dp/0394737873/.
Lund, Christian. 2014. “Of What Is This a Case?: Analytical Movements in Qualitative Social Science Research.” Human Organization 73 (3): 224–34. doi:10.17730/humo.73.3.e35q482014x033l4.