social_science   509

« earlier    

OSF Preprints | Explanation, prediction, and causality: Three sides of the same coin?
In this essay we make four interrelated points. First, we reiterate previous arguments (Kleinberg et al 2015) that forecasting problems are more common in social science than is often appreciated. From this observation it follows that social scientists should care about predictive accuracy in addition to unbiased or consistent estimation of causal relationships. Second, we argue that social scientists should be interested in prediction even if they have no interest in forecasting per se. Whether they do so explicitly or not, that is, causal claims necessarily make predictions; thus it is both fair and arguably useful to hold them accountable for the accuracy of the predictions they make. Third, we argue that prediction, used in either of the above two senses, is a useful metric for quantifying progress. Important differences between social science explanations and machine learning algorithms notwithstanding, social scientists can still learn from approaches like the Common Task Framework (CTF) which have successfully driven progress in certain fields of AI over the past 30 years (Donoho, 2015). Finally, we anticipate that as the predictive performance of forecasting models and explanations alike receives more attention, it will become clear that it is subject to some upper limit which lies well below deterministic accuracy for many applications of interest (Martin et al 2016). Characterizing the properties of complex social systems that lead to higher or lower predictive limits therefore poses an interesting challenge for computational social science.
social_science  prediction  explanation  forecasting  causality  philosophy_of_science  duncan.watts  for_friends  teaching 
4 days ago by rvenkat
We Are All Research Subjects Now - The Chronicle of Higher Education
These brakes on social inquiry are the same ones that academic researchers labor under — and sometime chafe under — today. And they are the same ones that the Social Data Initiative enlists in its public statements. But we should note that they were designed for research situations like the tearoom ethnography: where the privacy intrusion was intentional, where the potential harm to individuals’ dignity was obvious, where specific consent from the human subjects might feasibly have been obtained, and where careful scholarly review might have prompted a more ethical research design.

The data sets that Facebook plans to hand over to SSRC-approved researchers are by nature quite different. They first of all are being used only after the fact, having been collected via no peer-review process by a for-profit company. Accepting the terms of service of a social-media company is a far lower bar than the "informed consent" required by an IRB. These will be reams of personal data, possibly quite sensitive, and gathered unobtrusively, without the express consent (and often, knowledge) of those being researched.

It is not even certain whether the donors of data in this new venture are "research participants" in the sense that social scientists of the last century would have recognized. Some commentators have argued that because the company’s data will be anonymized before researchers get ahold of them (itself a concern as re-identification techniques improve), standards of informed consent do not even pertain.

Can the protections intended for a relatively small group of identifiable subjects in a bounded study — the men in St. Louis’s public restrooms in 1966, say — be extrapolated to the more than one billion virtual subjects who have been swept willy nilly into Facebook’s informational cache? The SSRC, in its early statements about the Social Data Initiative, seems to believe so. But today’s system of IRBs and federal regulations ought not be treated as definitive, especially given the new risks and possibilities presented by industry partnerships and big data.

Rather than accept the solutions of the 1960s and 1970s as a given, the new initiative would do better to reopen the questions that Tearoom Trade and other cutting-edge social research of its day generated about the legitimate bounds of social inquiry. The regulations that emerged were important, but so was the larger claim that human dignity ought to serve as an essential check on research ambitions....

For those who care both about pathbreaking social research and the rights of human subjects, the SSRC-Facebook collaboration poses dilemmas equivalent to those raised by Tearoom Trade. It is an opportunity to reconsider, and possibly revise, the rules of social inquiry. Are the guidelines for ethical research and treatment of subjects that were devised nearly 50 years ago a durable resource for us today? What kind of help can these tools, forged in quite different conditions, offer us in resolving the potential privacy violations and misuses of personal information that threaten today’s unwitting subjects of social media — and perhaps now scholarly — experimentation and manipulation?...

If we are all research subjects now, what kind of practices and policies will best preserve the values of individual dignity, privacy, and consent?

Given the unique nature of the new collaboration, these questions should be directed to the social scientists who will be making use of novel data sets. But they must also be answered by the corporations and data miners they collaborate with. If the byproduct is a new standard of data ethics with a broad purchase — viewed as the responsibility not simply of academics but also of the multifarious parties now engaged in social and behavioral research — that will truly fulfill the SSRC’s mission to "produce findings that improve everybody’s lives."
research  methodology  ethics  digital_methods  social_science  consent  IRB 
4 weeks ago by shannon_mattern
Nicolas Rashevsky's Mathematical Biophysics | SpringerLink
This paper explores the work of Nicolas Rashevsky, a Russian émigré theoretical physicist who developed a program in “mathematical biophysics” at the University of Chicago during the 1930s. Stressing the complexity of many biological phenomena, Rashevsky argued that the methods of theoretical physics – namely mathematics – were needed to “simplify” complex biological processes such as cell division and nerve conduction. A maverick of sorts, Rashevsky was a conspicuous figure in the biological community during the 1930s and early 1940s: he participated in several Cold Spring Harbor symposia and received several years of funding from the Rockefeller Foundation. However, in contrast to many other physicists who moved into biology, Rashevsky's work was almost entirely theoretical, and he eventually faced resistance to his mathematical methods. Through an examination of the conceptual, institutional, and scientific context of Rashevsky's work, this paper seeks to understand some of the reasons behind this resistance.
history  bio-physics  social_science  animal_behavior  mathematics 
july 2018 by rvenkat

« earlier    

related tags

?  abortion  academia  academic  addiction  administrative_data  administrative_state  ai  algorithms  amusing  analysis  andrew.gelman  angus.deaton  animal_behavior  anime  artificial_intelligence  asymmetric_information  bias  big_data  bio-physics  blog  business  cars  cass.sunstein  causality  chetty  children  christianity  commentary  computational_science  computer_science  consent  course  courses  critique  cs  css  culture  data  datasets  digital_methods  discussion  drugs  duncan.watts  economics  education  elite_opinion  entrepreneurship  epistemology  ethics  ethnography  evolutionary_psychology  experimental_design  experiments  explanation  facebook  fake_news  fb  film  flynn_effect  food  for_friends  forecasting  gafa  gaming  gender  genomics  gerrymandering  governance  government  guns  health  henry.farrell  history  hypocrisy  inefficiency  informative  infosec  insightful  institutions  interesting  internet  intervention  iq  irb  japan  law_enforcement  learning  lecture_notes  lgbt  machine_learning  marketing  masculinity  math  mathematics  measurement  medicine  mental_health  meso  meta-analysis  metaresearch  methodology  methods  modeling  moral_philosophy  nature-nurture  networked_public_sphere  networks  neuroscience  new_media  norms  nytimes  observational_studies  online_experiments  paper  philosophy  philosophy_of_science  philosophy_of_technology  policy  political_science  politics  pornography  prediction  privacy  pseudoscience  psychology  public_opinion  public_perception_of_science  public_policy  public_sphere  racism  regulation  religion  replication_of_studies  reproducible_research  research  research_practice  science  sex  signaling  social_epistemology  social_media  sociology  spatial_statistics  statistics  strength_training  structure  studies  study  teaching  to_read  us_supreme_court  usa  women's_studies  writing 

Copy this bookmark: