TY - JOUR
T1 - Navigating Statistical Uncertainty
T2 - How Urban and Regional Planners Understand and Work With American Community Survey (ACS) Data for Guiding Policy
AU - Jurjevich, Jason R.
AU - Griffin, Amy L.
AU - Spielman, Seth E.
AU - Folch, David C.
AU - Merrick, Meg
AU - Nagle, Nicholas N.
N1 - Funding Information:
This material is based upon work supported by the National Science Foundation under Grant Number 1132008. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
Funding Information:
This material is based upon work supported by the National Science Foundation under Grant Number 1132008. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. The authors thank the the editor and two anonymous reviewers, as well as Sy Adler, Charles Rynerson, Nick Chun, and Paul Lask for their feedback on earlier versions of this manuscript. We are also grateful to those planners participating in this research, including the Professional Development Officers of the American Planning Association chapters. The opinions, findings, and conclusions expressed in this publication are solely those of the authors and do not necessarily reflect the views of the supporting agency, and all errors and shortcomings are our own.
Publisher Copyright:
© 2018 American Planning Association, Chicago, IL.
PY - 2018/4/3
Y1 - 2018/4/3
N2 - Problem, research strategy, and findings: The American Community Survey (ACS) is a crucial source of sociodemographic data for planners. Since ACS data are estimates rather than actual counts, they contain a degree of statistical uncertainty—referred to as margin of error (MOE)—that planners must navigate when using these data. The statistical uncertainty is magnified when one is working with data for small areas or subgroups of the population or cross-tabulating demographic characteristics. We interviewed (n = 7) and surveyed (n = 200) planners and find that many do not understand the statistical uncertainty in ACS data, find it difficult to communicate statistical uncertainty to stakeholders, and avoid reporting MOEs altogether. These practices may conflict with planners’ ethical obligations under the AICP Code of Ethics to disclose information in a clear and direct way. Takeaway for practice: We argue that the planning academy should change its curriculum requirements and that the profession should improve professional development training to ensure planners understand data uncertainty and convey it to users. We suggest planners follow 5 guidelines when using ACS data: Report MOEs, indicate when they are not reporting MOEs, provide context for the level of statistical reliability, consider alternatives for reducing statistical uncertainty, and always conduct statistical tests when comparing ACS estimates.
AB - Problem, research strategy, and findings: The American Community Survey (ACS) is a crucial source of sociodemographic data for planners. Since ACS data are estimates rather than actual counts, they contain a degree of statistical uncertainty—referred to as margin of error (MOE)—that planners must navigate when using these data. The statistical uncertainty is magnified when one is working with data for small areas or subgroups of the population or cross-tabulating demographic characteristics. We interviewed (n = 7) and surveyed (n = 200) planners and find that many do not understand the statistical uncertainty in ACS data, find it difficult to communicate statistical uncertainty to stakeholders, and avoid reporting MOEs altogether. These practices may conflict with planners’ ethical obligations under the AICP Code of Ethics to disclose information in a clear and direct way. Takeaway for practice: We argue that the planning academy should change its curriculum requirements and that the profession should improve professional development training to ensure planners understand data uncertainty and convey it to users. We suggest planners follow 5 guidelines when using ACS data: Report MOEs, indicate when they are not reporting MOEs, provide context for the level of statistical reliability, consider alternatives for reducing statistical uncertainty, and always conduct statistical tests when comparing ACS estimates.
KW - American Community Survey (ACS)
KW - demographic data
KW - margin of error (MOE)
KW - statistical uncertainty
UR - http://www.scopus.com/inward/record.url?scp=85045088371&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85045088371&partnerID=8YFLogxK
U2 - 10.1080/01944363.2018.1440182
DO - 10.1080/01944363.2018.1440182
M3 - Article
AN - SCOPUS:85045088371
SN - 0194-4363
VL - 84
SP - 112
EP - 126
JO - Journal of the American Planning Association
JF - Journal of the American Planning Association
IS - 2
ER -