Person:

Lazer, David

In response to its poor performance during the 2012-2013 flu season, Google Flu Trends (GFT) engineers announced a redesign of the GFT algorithm. Two changes were made: (1) dampening anomalous media spikes and (2) using ElasticNet, rather than regression, for estimation. This paper identifies several problems that persist in the new algorithm. First, the transparency problems identified in our earlier Science paper appear to have, if anything, become worse. Second, there are reasons to doubt whether a spike in media attention was the only, or primary, cause of GFT's errors. Finally, there is strong evidence that GFT is still not using all the information at its disposal to make accurate measurements of flu prevalence. While it is too early to give a complete evaluation of the new algorithm, these results are discouraging.

The rise of fake news highlights the erosion of long-standing institutional bulwarks against misinformation in the internet age. Concern over the problem is global. However, much remains unknown regarding the vulnerabilities of individuals, institutions, and society to manipulations by malicious actors. A new system of safeguards is needed. Below, we discuss extant social and computer science research regarding belief in fake news and the mechanisms by which it spreads. Fake news has a long history, but we focus on unanswered scientific questions raised by the proliferation of its most recent, politically oriented incarnation. Beyond selected references in the text, suggested further reading can be found in the supplementary materials.

How do decentralized systems collectively solve problems? Here we explore the interplay among three canonical forms of collective organization—markets, networks, and hierarchies—in aggregating decentralized problem solving. We examine these constructs in the context of how the offices of members of Congress individually and collectively wrestle with the Internet, and, in particular, their use of official websites. Each office is simultaneously making decisions about how to utilize their website. These decisions are only partially independent, where offices are looking at each other for lessons, following the same directives from above about what to do with the websites, and confront the same array of potential vendors to produce their website. Here we present the initial results from interviews with 99 Congressional offices and related survey of 100 offices about their decisions regarding how to use official Member websites. Strikingly, we find that there are relatively few efforts by offices to evaluate what constituents want or like on their websites. Further, we find that diffusion occurs at the “tip of the iceberg”: offices often look at each others’ websites (which are publicly visible), but rarely talk to each other about their experiences or how they manage what is on their websites (which are not publicly visible). We also find that there are important market drivers of what is on websites, with the emergence of a small industry of companies seeking to serve the 440 Members. Hierarchical influences—through the House and through the party conferences—also constrain and subsidize certain practices.

A field is emerging that leverages the capacity to collect and analyze data at a scale that may reveal patterns of individual and group behaviors.

We surveyed 19,058 individuals across all 50 states plus the District of Columbia. The survey was conducted on 10-26 July 2020 by PureSpectrum via an online, nonprobability sample, with state-level representative quotas for race/ethnicity, age, and gender (for methodological details on the other waves, see covidstates.org). In addition to balancing on these dimensions, we reweighted our data using demographic characteristics to match the U.S. population with respect to race/ethnicity, age, gender, education, and living in urban, suburban, or rural areas. This was the seventh in a series of surveys we have been conducting since April 2020, examining attitudes and behaviors regarding COVID-19 in the United States.

We surveyed 19,058 individuals across all 50 states plus the District of Columbia. The survey was conducted on 10-26 July 2020 by PureSpectrum via an online, nonprobability sample, with state-level representative quotas for race/ethnicity, age, and gender (for methodological details on the other waves, see covidstates.org). In addition to balancing on these dimensions, we reweighted our data using demographic characteristics to match the U.S. population with respect to race/ethnicity, age, gender, education, and living in urban, suburban, or rural areas. This was the seventh in a series of surveys we have been conducting since April 2020, examining attitudes and behaviors regarding COVID-19 in the United States.

In June and July 2020, we surveyed 37,325 individuals across all 50 states plus the District of Columbia. The two survey waves used in this report were conducted on 12-28 June and 10-26 July 2020 by PureSpectrum via an online, nonprobability sample, with state-level representative quotas for race/ethnicity, age, and gender (for methodological details on the other waves, see covidstates.org). In addition to balancing on these dimensions, we reweighted our data using demographic characteristics to match the U.S. population with respect to race/ethnicity, age, gender, education, and living in urban, suburban, or rural areas. This was part of a project including a series of large-scale surveys we have been conducting since April 2020, examining attitudes and behaviors regarding COVID-19 in the United States.

We surveyed 19,052 individuals across all 50 states plus the District of Columbia. The survey was conducted on 10-26 July 2020 by PureSpectrum via an online, nonprobability sample, with state-level representative quotas for race/ethnicity, age, and gender (for methodological details on the other waves, see covidstates.org). In addition to balancing on these dimensions, we reweighted our data using demographic characteristics to match the U.S. population with respect to race/ethnicity, age, gender, education, and living in urban, suburban, or rural areas. This was the seventh in a series of surveys we have been conducting since April 2020, examining attitudes and behaviors regarding COVID-19 in the United States.

We surveyed 20,333 individuals across all 50 states plus the District of Columbia. The survey was conducted on 2-15 May 2020 by PureSpectrum via an online, nonprobability sample, with state-level representative quotas for race/ethnicity, age, and gender. In addition to balancing on these dimensions, we reweighted our data using demographic characteristics.

We surveyed 18,132 individuals across all 50 states plus the District of Columbia. The survey was conducted on 16-31 May 2020 by PureSpectrum via an online, nonprobability sample, with state-level representative quotas for race/ethnicity, age, and gender. In addition to balancing on these dimensions, we reweighted our data using demographic characteristics.