exam start at dec 13 7pm, 3 hours
Data C104 Final exam informational session Data C104: Human Contexts and Ethics of Data Final exam informational session We’ll start at Berkeley time (3:10) Plan for this informational session ● Format of the exam ● What it covers ● Key reading list ● How to study ● Sample Questions Exam Format Exam type: Multiple choice questions. Some questions may ask to you select “all that apply” Time: The exam will be 90 minutes long, and will take place Tuesday, December 13th, 7-8:30 Length: The exam will have between 25-40 multiple choice questions (we’ll specify the exact number after we’ve calibrated) Your exam questions will be randomly selected for each exam from a larger pool of questions The exam is open-note and open-book, but you may not discuss the exam with anyone else. You will be able to view all of your exam questions as soon as you begin the exam, and can change your answer to any question until the exam ends. What does it cover? The exam is comprehensive - it will cover material from the entire semester. But there will be a stronger emphasis on material since the midterm (units 4, 5, and 6) The exam covers lecture material (what’s on the slides and what was said in lecture) and associated readings. There will be questions about specific texts, but only the most important ones (listed on the next slide). We will not be asking “trivia” or questions about obscure facts. All questions about the readings will focus on their main arguments. Key texts Unit 4 ● James C. Scott, Seeing Like a State (selections) ● Emily Klancher Merchant, Building the Population Bomb ● Virginia Eubanks, “The Allegheny Algorithm” Unit 5 ● T. Porter, Trust in Numbers ● danah boyd, "You Think You Want Media Literacy: Do You?" ● Z. Tufekci, Twitter and Tear Gas Unit 6 ● Kate Crawford, "Labor," ● L. Vinsel and A. Russel, "Hail the maintainers” Unit 1 ● L. Winner, "Do Artifacts Have Politics?" ● Bowker and Star, "The Case of Race Classification and Reclassification Under Apartheid” ● R. Benjamin, Race After Technology Ch. 3 “Coded Exposure” Unit 2 ● Winner, "Brandy, Cigars, and Human Values” ● Amia Srinivasan, "Stop the Robot Apocalypse," ● Madeleine Clare Elish, "Moral Crumple Zones” Unit 3 ● J. Radin, "Digital Natives" ● Dwork and Mulligan (2013) "It’s not privacy, and it’s not fair. ● C. Koopman, How We Became Our Data - “Introduction” Additional information Exam support: Professor Carson and Professor Edmundson will be available by email and (likely) by zoom during the exam to answer any questions. We’ll send out more details as we know more Alternative exam window: For those of you with exam conflicts, we will schedule an altenative exam window for Monday, December 12th or earlier in the day on Tuesday, December 13th. Please fill out the survey posted to bCourses. We will provide details soon. How to study This exam is meant to assess your understanding of course materials in a similar way to the midterm exam. It focuses more on conceptual understanding than on memorizing facts (though some factual knowledge is obviously essential!). The study process should be similar to the midterm! - Review key terms and tools - Review lecture recordings for major arguments and take notes, or review your lecture notes. - Review your reading notes on key readings Sample question #1 The Belmont Report was written in response to: A. Three Mile Island B. Challenger Space Shuttle disaster C. Tuskegee Syphilis Experiment D. Cambridge Analytica/Facebook scandal E. The Pima Indians Diabetes Database (PIDD) Sample Question #2 Which of the following are the four technological “layers” of the datafied world, ordered from historically oldest to most recent? A. Quantification, Computerization, Machine Learning, Artificial Intelligence B. Ubiquitous data, Reliable Analytics, Culture of Algorithms, Autonomous Agents C. Data Collection Imperative, Feedback loops, Algorithmic Decision-making, Intelligent Machines D. Autonomous Agents, Reliable Analytics, Culture of Algorithms, Ubiquitous data Sample Question #3 Which of the following most clearly articulates a major analytical conclusion we drew from the COMPAS case study in class? Choose the single best answer. A. COMPAS is a racist and unfair algorithm because it reflects the racial biases of its designers B. Algorithms used for pre-trial detention are fair because they make more neutral and objective decisions than human beings C. The concept of “fairness” is meaningless as a metric for evaluating algorithms because it has many competing definitions D. Claims about the fairness of risk assessment algorithms should be evaluated not only in terms of the algorithm’s outputs, but in terms of the specific sociotechnical contexts in which they are deployed. Sample question #4 Which of the following best captures the meaning of the claim that “personal data is a sociotechnical concept”? A. “Personal data” is sociotechnical because it is information collected by people about people using technological systems like computers, and so all personal data has both social and technical components. B. “Personal data” is a sociotechnical concept because what counts as “personal data” is historically contingent and relational: it depends on the technical infrastructures used to collect the data, the laws and norms that govern data collection practices, and the purposes for which that data is used. C. “Personal data” is a sociotechnical concept because the systems that collect data about people are complex and redistribute risk and responsibility widely and unevenly, posing unique challenges for regulation and intervention.