44. Putting on Something More Revealing: Exposing Our Fedora Data (41:37)

Date:
2014-04-02
Main contributors:
Hardesty, Julie, Shelby, Jacob
Summary:
IU's Fedora digital repository houses thousands of digitized items including pages of text, photographs, puzzles, three-dimensional artifacts, prints, audio, and moving pictures. These items are made accessible either through silo-ed collection sites or as a certain type of digital object (like images or finding aids). But there is a larger corpus of the entire repository that should be shared for greater understanding, discoverability, and use. Exposing the metadata we have in our repository allows others to make use of it, offering different perspectives on our items and collections and combining our collections with other similar collections around the world. Join us for the latest on our work to reveal our digital collections as data feeds.

45. Simple Heuristics for a Complex World (1:43:32)

Date:
2014-03-27
Main contributors:
Gerd Gigerenzer
Summary:
Whom to marry? How to invest? Whom to trust? Complex problems require complex solutions – so we might think. And if the solution doesn’t work, we make it more complex. That recipe is perfect for a world of known risks, but not for an uncertain world, as the failure of the complex forecasting methods leading to the 2008 financial crisis illustrates. In order to reduce estimation error, good inferences under uncertainty counter-intuitively require ignoring part of the available information,. Less can be more. Yet although we face high degrees of uncertainty on a daily basis, most of economics and cognitive science deals exclusively with lotteries and similar situations in which all risks are perfectly known or can be easily estimated. In this talk, I invite you to explore the land of uncertainty, where mathematical probability is of limited value and people rely instead on simple heuristics, that is, on rules of thumb. We meet Homo heuristicus, who has been disparaged by many psychologists as irrational for ignoring information—unlike the more diligent Homo economicus. In an uncertain world, however, simple heuristics can be a smart tool and lead to even better decisions than with what are considered rational strategies. The study of heuristics has three goals. The first is descriptive: to analyze the heuristics in the “adaptive toolbox” of an individual or an institution. The second goal is normative: to identify the ecological rationality of a given heuristic, that is, the structures of environments in which it succeeds and fails. The third goal is engineering: to design intuitive heuristics such as fast-and-frugal trees that help physicians make better decisions.

49. The Failure of Altmetrics (58:55)

Date:
2014-11-12
Main contributors:
Sugimoto, Cassidy
Summary:
In this talk, Cassidy Sugimoto argues that altmetrics have failed to deliver on their promise. She discusses criticisms of altmetrics (including those dealing with validity and reliability issues), but argues that the largest failure of altmetrics has been the focus on a single genre‰ÛÓthat is, the journal article‰ÛÓand setting altmetrics up as an alternative to citations. Sugimoto introduces the notion of outcomes-based evaluation and demonstrates that altmetrics cannot be equated with outcomes in this model. She urges the community to rethink ways in which we can build metrics that can capture larger societal impact. She discusses four axes of potential impact: production, dissemination, engagement, assessment. In each of these, she reviews various examples of current initiatives and challenges the audience to conceive of possible metrics to capture the desired outcome in each scenario.