How SENTRY Works
SENTRY measures a potential respondent’s capacity to adequately participate in a survey at a given time, through a series of behavioral measurements proven to promote respondents who are attentive, comprehending, engaged and respond consistently to survey tasks. SENTRY also manages participants through device fingerprinting to prevent duplication and fraud. This process takes less than a minute and reduces back-end data cleaning by removing unengaged and fraudulent respondents before they enter a survey.
How SENTRY Helps
Good data leads to good decisions. Noise can obscure truth, and lead to very inaccurate conclusions. We recently analyzed a report by the CDC and found that their conclusions appear to be due to this kind of noise. As you can see below, SENTRY eliminates this noise.
In the past month, have you or a household member engaged in the following behaviors to prevent coronavirus?
Very high-risk cleaning practices were almost exclusively reported by respondents who did not pass SENTRY. This indicates that reports of very high-risk cleaning practices are largely due to noise.
Better Open-Ended Responses
95% of respondents who pass SENTRY provide quality open-ended data.
Sample Response Accuracy
SENTRY only allows participants who are attentive and truthful into your survey. Below you can see ‘Red Herring’ questions that respondents who pass SENTRY are very likely to get correct, while those who fail SENTRY are likely to answer incorrectly.
Traditional ‘Red Herring’ questions and other attention checks, disproportionately remove minority participants and those with lower education. SENTRY preserves the diversity of these hard-to-recruit groups. The graph below shows that while Socio-economically disadvantaged groups are much more likely to fail some traditional attention checks, SENTRY is not biased against them.