class: center, middle, inverse, title-slide # Incorporating Data Ethics In a Statistics/Data Science Major ## (or at least starting to) ### 🖳Miles Ott
Smith College Program in Statistical & Data Sciences
### SDSS Bellevue Washington, May 31, 2019 --- # Assuming everyone agrees that Data Ethics is important -- # Smith SDS in the __*process*__ of addressing how to incorporate Data Ethics into the program / major --- class: center, middle #Background on Smith College Statistical and Data Sciences program --- # SDS Tenure-track Faculty .pull-left[ 100% in SDS ![:scalewidth 33%](https://www.smith.edu/sites/default/files/styles/img-faculty-detail/public/media/Faculty/ben_baumer_crop.jpg) Ben Baumer ![:scalewidth 33%](https://www.smith.edu/sites/default/files/styles/img-faculty-detail/public/media/Faculty/miles_ott_crop.jpg) Miles Ott ![:scalewidth 33%](https://www.smith.edu/sites/default/files/styles/img-faculty-detail/public/media/Faculty/albert_kim_crop.jpg) Albert Y. Kim ] .pull-right[ 50% in SDS ![:scalewidth 33%](http://ds.cs.umass.edu/sites/default/files/styles/yellow_pages/public/Photos/Profiles/halvorsen_130.jpg) Katherine Halvorsen (Math) ![:scalewidth 33%](https://www.smith.edu/sites/default/files/styles/img-faculty-detail/public/media/Faculty/randi_garcia_crop.jpg) Randi Garcia (Psychology) ![:scalewidth 33%](https://www.smith.edu/sites/default/files/styles/img-faculty-detail/public/media/Faculty/katherine_kinnaird.jpg) Katherine Kinnaird (Computer Science) ] --- # [Departmental Learning Goals](https://www.smith.edu/about-smith/institutional-research/learning-goals) -- ### Newly Adopted Ethics Learning Goal in Spring 2019: -- ### _Assess the ethical implications to society of data-based research, analyses, and technology in an informed manner. Use resources, such as professional guidelines, institutional review boards, and published research, to inform ethical responsibilities._ --- # [Departmental Learning Goals](https://www.smith.edu/about-smith/institutional-research/learning-goals) -- ![](slack_dept_learning_goals.GIF)<!-- --> -- ### Implementation of of this goal? -- ### Assessment of this goal? --- ### SDS Major Diagram <img src="sds_major_diagram_720px.png" width="1280" /> --- # Data Ethics Class? -- - ### Rather than have one course exclusively focused on data ethics... -- - ### __*Incorporate data ethics in many classes so that it is fully integrated throughout the curriculum*__ -- - ### Rather than having one day of class module on ethics... -- - ### __*Incorporate data ethics throughout the course*__ --- # Instead of _Patching_ Data Ethics into your course -- <img src="Data_ethics_patch.png" width="1707" /> --- # _Knit_ Data Ethics into the fabric of your course -- <img src="knit.jpg" width="660" /> --- class: inverse, center, middle # Example: Statistical Analysis of Social Network Data --- # Teaching a Class on Social Networks is Wicked Fun -- - ### Introducing a new kind of data -- - ### Clean slate: permission to be creative -- - ### Implications for Data Ethics: analysis, informed consent, data privacy, data sharing -- - #### Text Book: [Statistical Analysis of Social Network Data with R](https://link.springer.com/book/10.1007/978-1-4939-0983-4) by Eric D. Kolaczyk and Gábor Csárdi --- # First Day of Class -- <img src="Day_1_Class.png" width="1707" /> --- # First Homework Assignment: Introducing Network Data -- - ### Read [Ethical Considerations for Data Collection Using Surveys](https://onf.ons.org/file/30081/download). How does this reading relate to social network analysis? Did collecting the Grey's Anatomy data bring up ethical concerns? Why or why not? - #### Read pages 1-10 of the text, Summarize (in your own words) the three types of Network Analyses. How could each type help us answer questions about Grey’s Anatomy? - #### Read pages 13-18 of the text, then use the social network data you collected in class on Grey’s Anatomy first six minutes to make (by hand): An adjacency list, an edge list, and an adjacency matrix, etc --- # First Homework Assignment: Introducing Network Data -- - ### __Goal__: Students make their own connections between data ethics and social network data -- - ### __Goal__: Prime students to consider ethical implications _every time_ they encounter a new data set -- - ### __Goal__: Set the expectation that I will not be giving them answers on what is ethical and what isn't --- # Second Homework Assignment: Data Visualization Read the paper on [Zachary's Karate Club](https://www.jstor.org/stable/3629752?seq=1#page_scan_tab_contents) and reflect on the following questions: - __How does the way in which the data were collected inform your interpretation of the network?__ - __What are the ethical considerations for this research, and how were they addressed or not addressed?__ - Reflect on the network visualization in the paper. - Why do you think they chose to visualize the network in that way? - In what ways was the visualization effective or not effective? - If you were to analyze this network, what would you do differently? --- # A few homework assignments where ethics isn't explictly mentioned - Hypothesis Tests With Networks - Mathematical Models for Networks - Small World Networks - Random Graph Models - Exponential Random Graph Models (ERGM) - Stochastic Block Models (SBM) --- # Sixth Homework Assignment: Link prediction - Read pages 111-116 of Chapter 7 - Read this paper [The Link Prediction Problem for Social Networks](https://www.cs.cornell.edu/home/kleinber/link-pred.pdf) - Search to find readings (blog posts are OK!) -- - What are important qualities in a link prediction algorithm? - What are unusual settings for social network link prediction (beyond what we spoke about in class)? - __What are ethical considerations for link prediction?__ - __What is the most unethical link prediction situation that you can think of? Feel free to be creative.__ - __What would be a link prediction situation that could positively benefit society?__ - How can companies use link prediction for monetary gain? - What are some of the challenges when trying to predict links? - What questions do you have about link prediction? --- #Seventh Homework: Social Networks and Data Ethics - Read chapter 4 of [Gathering Social Network Data](https://us.sagepub.com/en-us/nam/gathering-social-network-data/book260973) - Read ["I Didn't Sign Up for This!": Informed consent for social network research](https://www.aaai.org/ocs/index.php/ICWSM/ICWSM15/paper/viewFile/10493/10501) - Search to find readings (blog posts are OK!) about ethics and social network analysis -- - What are some ethical considerations for social networks you hadn't considered before? - What are some of the biggest risks from social network analysis? - What are some of the biggest benefits from social network analysis? - Do you think your biggest risk would be offset by your biggest benefit? - What are some things to avoid when presenting network results? - What would you want to include in your social network code of ethics? - Explain the main point of the paper on informed consent. - What are two or three new concepts from the paper on informed consent? - What questions came up for you in any of your readings? --- #The Rest of the Semester - ### Causal Inference for Social Networks - ### Public Health and Social Networks - ### Network Sampling Methods (Respondent-Driven Sampling) -- - ### Final Group Project: Network Data Analysis --- # Some Opinions / Suggestions: -- - ### Ethics Integrated throughout >> Stand alone ethics modules >> no mention of data ethics at all -- - ### Statistical and Data science students can love thinking about data ethics -- - ### There are many ways to knit Data Ethics into program / department (Learning goals, hiring new faculty, designing classes, capstone projects, etc) -- - ### You can do this! --- class: inverse, center, middle # Thank you! <i class="fa fa-envelope" aria-hidden="true"></i> [mott@smith.edu](mailto:mott@smith.edu) <i class="fa fa-twitter" style="font-size:36px"></i> [Miles_Ott](http://www.twitter.com/Miles_Ott) <i class="fa fa-github" style="font-size:36px"></i> [MilesOtt](http://www.github.com/MilesOtt) Slides Available here: [https://bit.ly/2QBKTg1](https://bit.ly/2QBKTg1) --- # Hiring -- - ### Data ethics statement -- - ###Address data ethics in teaching and/or research statement -- - ###Address data ethics during the early interview phase -- - ###Teaching demonstration to incorporate data ethics