Consent Form Mechanical Turk: Fill & Download for Free

No, this isn’t a deal. (Not anymore, at least.)First off: Cambridge Analytica did not harvest Facebook data.They paid someone else to give them psychographic profiles of the US voter base. That someone else was Dr. Aleksandr Kogan, a lecturer at Cambridge University.How did Kogan harvest all that Facebook data?The simple answer is that Facebook let him.Until 2014, FB users could share data from their friends lists with app makers. Remember when an app would request access to your profile and friends list?(Illustrations from The Facebook/Cambridge Analytica Data Scandal, Visually Explained)You were giving that app access to your friends’ profile data. Age, gender, religious affiliations, likes, interests, etc.Which is weird, because while everyone knows THIS is creepy:THIS was totally OK on Facebook for seven years.The problem here: individuals could “give up” their friends’ data without those friends’ consent. Facebook shut this feature down in 2014.But not before Kogan collected profile data on about 87 million FB users.[1][1][1][1]How did Kogan get those original 300,000 people to take the personality quiz?Easy: he paid them to.Amazon’s Mechanical Turk is a huge pool of human task-doers. Kogan paid 270k–300k people to take the personality quiz at about $3–4 USD a pop. Those people allowed the app to access their friends lists, and—voilà!—Kogan had data on 87 million people!Now, if you ask me, that’s a deal. But it only worked because Facebook allowed apps to ask users for their friends’ data—without their friends’ consent. That’s not allowed anymore.Footnotes[1] Facebook Says Cambridge Analytica Harvested Data of Up to 87 Million Users[1] Facebook Says Cambridge Analytica Harvested Data of Up to 87 Million Users[1] Facebook Says Cambridge Analytica Harvested Data of Up to 87 Million Users[1] Facebook Says Cambridge Analytica Harvested Data of Up to 87 Million Users

I’ve been dubious about these ‘quizzes’ for a while.When looking at the sites behind the flavour of the day ‘what political/personality type are you’ doing the rounds on platforms like Facebook, I found a number of them had properties that were concerning.Information on the company behind them is simply a (copyright)website name. No other information.Any research type content is a cut and paste type quality.They claim to require demographic information to be taken from a person’s profile in order to give a ‘result’. Which makes no sense.They do tend to state that ‘they will not share your data with third parties’ and they say that individual data will be used in order to ‘improve’ their ‘quizzes’. Which suggests model building.Do they value the ‘result’, or are they more interested in the ‘answers’ to the ‘quiz questions’ combined with any demographic individual data they acquire as a person takes the ‘test’?The Cambridge Analytica files were definitely interesting in that regard.‘Kogan then set up GSR to do the work, and proposed to Wylie they use the data to set up an interdisciplinary institute working across the social sciences. “What happened to that idea,” I ask Wylie. “It never happened. I don’t know why. That’s one of the things that upsets me the most.”It was Bannon’s interest in culture as war that ignited Wylie’s intellectual concept. But it was Robert Mercer’s millions that created a firestorm. Kogan was able to throw money at the hard problem of acquiring personal data: he advertised for people who were willing to be paid to take a personality quiz on Amazon’s Mechanical Turk and Qualtrics. At the end of which Kogan’s app, called thisismydigitallife, gave him permission to access their Facebook profiles. And not just theirs, but their friends’ too. On average, each “seeder” – the people who had taken the personality test, around 320,000 in total – unwittingly gave access to at least 160 other people’s profiles, none of whom would have known or had reason to suspect.What the email correspondence between Cambridge Analytica employees and Kogan shows is that Kogan had collected millions of profiles in a matter of weeks. But neither Wylie nor anyone else at Cambridge Analytica had checked that it was legal. It certainly wasn’t authorised. Kogan did have permission to pull Facebook data, but for academic purposes only. What’s more, under British data protection laws, it’s illegal for personal data to be sold to a third party without consent.“Facebook could see it was happening,” says Wylie. “Their security protocols were triggered because Kogan’s apps were pulling this enormous amount of data, but apparently Kogan told them it was for academic use. So they were like, ‘Fine’.”Kogan maintains that everything he did was legal and he had a “close working relationship” with Facebook, which had granted him permission for his apps.Cambridge Analytica had its data. This was the foundation of everything it did next – how it extracted psychological insights from the “seeders” and then built an algorithm to profile millions more’ - ‘I made Steve Bannon’s psychological warfare tool’: meet the data war whistleblowerThere are a number of questions about these tests on Quora -what is your result on test X - type thing. It’d be interesting to look back at them forensically.

I added this question a couple of months ago, when I was reading Salganik’s book and had just learned of Gónzalez-Bailón’s through Twitter. Having finished Decoding the Social World last night, I can now say that, despite covering some similar ground, these are substantively different books which serve different purposes.Before getting into how the two books differ, I’ll note that both are motivated by the same historical trend: the proliferation over the past decade, primarily through the worldwide adoption of web products, of data relevant to the social sciences. This has led to the birth of a new field, called “computational social science.” Computational social science is of course of interest to social scientists, who now have the ability to quantitatively test models and theories on an unprecedented scale. However, the field has also attracted people from across the sciences, who (like me!) do not have a background in sociology.Salganik’s Bit by Bit is a practical guide for bridging the divides present at the birth of computational social science. For people from a traditional background in the social sciences, this book covers the ways in which online data differs from traditional sociological data, the opportunities and challenges that arise from running surveys in the context of web products, how to think about data that is crowdsourced through things like Mechanical Turk, etc. For people from other fields, Bit by Bit covers ideas from the social sciences that remain important in the “digital age.” For both groups, the book discusses milestones from the first decade of computational social science. Bit by Bit is very clearly and accessibly written, and I’m pretty confident that it will be a classic reference for the field.On the other hand, Gónzalez-Bailón’s book places modern computational social science in the context of the long-sweep of the history of sociology. Decoding the Social World is not a textbook, and you can’t learn how to be a practitioner of computational social science by reading it. Instead, I think the value of the book is more inspirational, because reading it gives a sense of how modern computational social science fits into a grand intellectual tradition. Gónzalez-Bailón argues throughout the book that the problems raised by modern communication platforms have precedents in the problems raised by communication revolutions over the centuries. She also surveys how sociologists of the past have identified many of the problems that we think about today, with a particular emphasis on the debates between Émile Durkheim and Gabriel Tarde in late 19th Century France. The main difference of our era, she argues, is that we can log and study data at unprecedented scale, and this can help shed light on sociological issues that recur with each communication revolution.Incidentally, Gónzalez-Bailón’s book also contains my favorite articulation of what “computational social science” is, which I’ll share here (from page 52 of Decoding the Social World):Today, digital technologies have placed the analysis of spontaneous and unstructured social action (with its long theoretical lineage) back at the heart of social inquiry, enabling a research agenda that has come to be known as computational social science… Insights are now coming not just from the social sciences (with their historical boundaries) but from the many attempts to analyze the available evidence with new tools—new, at least, to how the proponents of collective behavior research traditionally pursued their work.As I read these two books, I added questions to Quora that were inspired by what I was reading. I’ve linked them all below. If you’re interested in reading either of them, maybe these questions can give you a better sense about their contents…Bit by BitWhat are some great examples from the research literature of "nowcasting," which Matthew Salganik defines in Bit by Bit as the "use [of] ideas from forecasting to measure the current state of the world?"What social science "natural-experiment" studies have been enabled by the birthday-based drafting system used by the United States during the Vietnam War?What techniques were invented to minimize bias in conducting surveys using random digit dialing?What is the Total Survey Error Framework, and how is it used to improve the statistical validity of survey analyses?What are some examples of survey experiments that revealed surprising question wording or question form effects (i.e., where semantically identical survey questions elicited very different responses)?What are the best practices for discovering and mitigating the biases introduced by question form and question wording effects in survey research?What are some examples of especially effective uses of the wiki survey framework developed by Matthew Salganik and Karen Levy?Why do people tend to systematically underestimate their level of disagreement with their friends? Is it because they make assumptions about others' perspectives, because they avoid conversations around controversial topics, or some other reason?Are there general trends to when social science studies extrapolate well or poorly from WEIRD (Western, Educated, Industrialized, Rich, and Democratic) study participants to more general demographics?Are there important confounders that are introduced by creating a social-science experiment that the participants enjoy too much compared to the real-life situations that the experiments are supposed to describe?What does "construct validity" mean in the context of a social science study, and how can we establish that construct validity holds?What are some examples of strong demand or Hawthorne effects in social-science experiments?What is a "mixed-design" experiment in the social sciences?In chapter 4 of Bit by Bit, Matthew Salganik implies that early concerns about experiments involving Amazon Mechanical Turk have proven to be unjustified: how was this established?What are the techniques that ornithologists use to mitigate concerns around biased sampling and data quality when using data from eBird?When online mass collaboration leads to a scientific publication, what are the principles that should be used in determining who gets credit as an author of the publication?In chapter 5 of Bit by Bit, Matthew Salganik suggests that one application of online mass collaboration would be an open call to identify the first time a person or idea was mentioned in writing: has something like this been done successfully?How are the prediction markets used to estimate the reproducibility of scientific research?What are some examples of cases where the non-research member of an Institutional Review Board contributed critical and very distinctive input into a review process?What are the "five safes" of a sound data protection plan (as described by the UK Data Services), and how do they collectively decrease informational risk?What are "context-relevant informational norms," and how should they guide the development of a study in the social sciences?What are some examples of social-science studies that have been abandoned because of the results of an "ethical-response survey" (as described by Salganik in Chapter 6 of Bit by Bit)?How can researchers ethically deal with the "transparency paradox" in informed consent during a social-science study?Decoding the Social WorldWhat did Auguste Comte and Adolphe Quételet mean when they used the term "social physics" in the 1830s?What are some empirically demonstrated examples of the "reflexivity problem" in social theories (where, in the attempt to understand social behavior, we change what society believes and how people behave)?What are the most important papers that study phase transitions in Schelling's segregation model and in variants of that model?What are the important similarities and differences in the perspectives of Émile Durkheim and Gabriel Tarde on sociology?How did the Dreyfus Affair impact the development of sociology in the late 19th century France?What is the two-step flow theory of communication of Lazarsfeld, and how has it been supported or challenged by modern social communication data?What should modern computational social scientists know about the social theories of Gabriel Tarde?What should modern computational social scientists know about the social theories of Émile Durkheim?What is the latest attested occurrence of a “dancing mania” of the type that occurred in Strasbourg in 1518? Why has it not recurred more recently?Are there attested cases of "dancing mania" from outside Europe? If not, then why was the phenomenon geographically confined to Europe?Did the urban layout of Leipzig help catalyze the Monday Demonstrations of 1989 in East Germany?What are Herbert Blumer's "elementary mechanisms of collective behavior," and do they still serve as a good categorization in the age of online social media?What is the "versatility" of a node in a multilayer network, and what is it useful for?How has the Meeting Mediator affected the dynamics of meetings that you have attended?