Using Auxiliary Population Samples for Sample-Selection Correction in Models Based on Crowd-Sourced Volunteered Geographic Information

Abstract

Citizen science (CS) projects (and some social media) offer selected samples with extensive information about human interactions with the natural world. We independently elicit levels of engagement with the eBird project from (1) members of the eBird CS project and (2) a general population sample. The general-population sample allows an ordered-probit model to explain propensities to engage with eBird, which we transfer to predict selection-correction terms for our independent sample of eBird members. We illustrate, using a question posed only to our eBird member survey sample about the radii of their individual spatial consideration sets for typical one-day birding excursions.Appendix materials can be accessed online at: https://uwpress.wisc.edu/journals/pdfs/LE-98-1-01-Cameron-app.pdf.

Citizen scientists: Unveiling motivations and characteristics influencing initial and sustained participation in an agricultural project

“…For example, in environmental valuation surveys two-step models correct for higher environmental values and/or concerns of self-selected survey respondents [ 76 , 77 ]. Two-step methods have also previously been applied in both volunteering and citizen science specifically, given that the decision to volunteer is made by individuals, indicating those who choose to volunteer constitute a self-selected sample rather than a random sample [ 71 , 72 ]. In our study, the significance of rho and the Likelihood Ratio tests validating the suitability of using two-step models.…”

“…As touched upon in the theoretical background, environmental citizen scientists display specific motivations and personal traits [ 13 , 34 , 37 , 43 ]. By voluntarily applying to join a citizen science project, participants effectively self-select themselves into the pool of citizen scientists, potentially leading to biases in estimation results of the outcome variables [ 71 , 72 ].…”

Explicit incentives increase citizen science recordings

“…Citizen science data have been used to document global processes such as biodiversity loss (Eichenberg et al, 2021), land-use change (Liu et al, 2022), invasive (Bowser et al, 2020;Burgess et al, 2017;Cohn, 2008). In principle, one can leverage the potentially large number of observations to factor out measurement errors, and one can correct for potential bias when its direction and size can be estimated (Bird et al, 2018;Cameron & Kolstoe, 2022;van Strien et al, 2013). For both tasks, it is essential to understand the incentives and motivations of the people that produce the data (Arazy & Malkinson, 2021;Bird et al, 2018;Kelling et al, 2015;van Strien et al, 2013).…”

Hypothetical bias mitigation in representative and convenience samples

“…For example, commodity boards may be unwilling to share individual member contact information. Recreational participants may not reveal themselves as part of the population via membership in formal organizations such as birding (Cameron & Kolstoe, 2022).…”