Student or educator? Sign up with your school email and get 6 months of Pro free.
Cited by 631

The DOI matches its Crossref record; 5 of the 5 references it cites were found in a registry or an index.

Big Data: A Revolution That Will Transform How We Live, Work, and Think

Naimi, A. I., Westreich, D. J.·2014·American Journal of Epidemiology

Abstract

“Big data” have big potential to influence society and science, including the way we approach and conduct epidemiologic research. In this book, Mayer-Schönberger and Cukier (1) outline 3 perspectives (or “shifts”) which are (the authors argue) inherent in big data: 1) the obsolescence of sampling (chapter 2, “More”); 2) the acceptance of increased measurement error in return for more data (chapter 3, “Messy”); and 3) a “move away from the age-old search for causality” (chapter 4, “Correlation”) (1, p. 14). The book is divided into 10 chapters. Chapter 5 discusses the move in industry and government to collect data on everything (including health-related topics); chapter 6 discusses the value of such data. Chapters 7–9 discuss the implications and risks of this practice and ways to control the “datafication” of human life. Chapter 10 summarizes future possibilities. In this review, we focus on chapters 2–4, which explore the 3 shifts which the authors claim big data will bring to the scientific table and which we believe will be of most interest (and most concern) to readers of the Journal. The chapter entitled “More” (chapter 2) explains how big data is defined by collecting as much data as possible, and at times “all” of it. The authors argue that new technologies will allow data scientists to passively collect, store, and analyze much more data in real time. In many instances, the authors refer to sampling as an outdated hindrance to discovery: “Reaching for a random sample in the age of big data is like clutching at a horse whip in the era of the motor car. … Increasingly, we will aim to go for it all” (1, p. 31). However, the authors seem to misunderstand random sampling. Describing an analysis in which “all the data” were used (in this case, every sumo match over 11 years) to discover a surprising pattern of match-throwing, the authors claim that “random sampling of the bouts might have failed to reveal it” because “without knowing what to look for, one would have no idea what sample to use” (1, p. 29). As far as we can understand with regard to this passage, the authors are confusing issues of precision and validity: A simple random sample of the 11 years' worth of sumo bouts might have had less power to reveal the finding than an analysis of all bouts, but it is nonsense to suggest that the same pattern would not have emerged in a simple random sample of the bouts. Of course, if it is possible and straightforward to do so, using all of the data is ideal; yet the perspective taken by the authors ignores the fact that a sample is defined relative to a target population (2), and we may not be able to collect data on “n = all” because of how we define our target. Relatedly, the authors fail to appreciate that their examples of “all the data” are frequently no such thing. Google Flu Trends (Google, Inc., Mountain View, California)—a frequent example in these chapters—uses a great deal of data, but it is data obtained from a subset of literate and (most likely) high-socioeconomic-status individuals. If we were interested in a disease that was more concentrated among persons without access to computers, would Google searches work as well? Similar criticisms can be made of the authors' claim that when the data is [sic] collected passively while people do what they normally do anyway, the old biases associated with sampling … disappear. We can now collect information that we couldn't before, be it relationships revealed via mobile phone calls or sentiments unveiled through tweets (1, p. 30). The idea that information revealed through tweets is free from bias is simply risible: Twitter (Twitter Inc., San Francisco, California) is used by perhaps 16% of Americans (3). Moreover, public health science is frequently faced with finite resources: It is far from clear that it is in the interests of public health to devote resources to (for example) expensive assays in a very large sample of individuals when a simple random sample will provide the same point estimate and negligibly worse precision. Chapter 3 (“Messy”) discusses measurement error, an inherent part of any data science and a central concern in epidemiology (4). The authors claim that an embrace of big data allows for greater acceptance of measurement error, and argue that “[i]n return for relaxing the standards of allowable errors, one can get ahold of much more data, [and] sometimes ‘more trumps better’ ” (1, p. 33). While the authors parenthetically mention a caveat about the need to avoid “systematic bias” (1, p. 34), they fail to explain that merely having greater amounts of systematically biased data will often lead to more problems rather than fewer. A meta-analysis of 10 studies with the same measurement error bias will be more precise, but no more valid, than any single one of those studies. At the same time, the increased precision will offer increased (but false) confidence in those fundamentally flawed results. It is not until much later in th

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/aje/kwu085

Cites

5 of 5 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 5 references

Metadata

Authors
Naimi, A. I., & Westreich, D. J.
Venue
American Journal of Epidemiology
Published
2014-04-08
Volume
179
Issue
9
Pages
1143-1144
ISSN
  • 0002-9262 (ISSN-L)
  • 1476-6256
OpenAlex
W2101468243
Cited by
631
References
5
FWCI
105.89
Type
journal article
Language
en
Open access
closed

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Big Data: A Revolution That Will Transform How We Live, Work, and Think. https://app.cochat.ai/papers/10.1093/aje/kwu085
The paper
Naimi, A. I., & Westreich, D. J. (2014). Big Data: A Revolution That Will Transform How We Live, Work, and Think. American Journal of Epidemiology. https://doi.org/10.1093/aje/kwu085

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.