1Cademy - Two Truths and a Lie: Exploring Soft Moderation of COVID-19 Misinformation with Amazon Alexa

Learn Before

Two Truths and a Lie: Exploring Soft Moderation of COVID-19 Misinformation with Amazon Alexa

Concept

Two Truths and a Lie: Exploring Soft Moderation of COVID-19 Misinformation with Amazon Alexa - Methodology

After users ask Alexa to access the Twitter Reader and Alexa receives a confirmation from the user to read their Twitter stream, Alexa gets all the tweets with warning labels and starts reading them, only focusing on plain text and ignoring symbols and characters such as the hashtag
In the first case, Alexa informs users of the warning label and asks them if they would like Alexa to keep reading the tweet. If a user answers no, Alexa will ask to read the next tweet in the stream, otherwise Alexa ends the conversation and exchange. The study on the accuracy of tweets compared user perceptions regarding tweets with vs. without warnings and Alexa mentioning the warning before vs. after reading the tweet contents
In the second case, Alexa simply reads the warning label after reading the whole tweet. The study on hesitancy and political leanings focused on whether a user’s political affiliations will impact their perceptions of the accuracy of the tweet contents through Alexa’s verbal warnings and reading
Participants recruited from Amazon Mechanical Turk and had to be at least 18 years old, uses Twitter and Alexa, and has seen at least one tweet related to COVID-19 vaccines
Experiment consists of participants answering questions – one about perceived accuracy where each person rated the content on a Likert scale (not accurate at all, slightly inaccurate, slightly accurate, or very accurate), a second on their intent to get the vaccine (yes, no, or I don’t’ know), and a final one about their political affiliation (Democrat, Republican, or Independent)
Participants were randomly assigned into one of 6 groups – repeated tweet with warning; repeated tweet without warning; warning before repeated tweet; repeated verified tweet; repeated verified tweet with warning; warning before repeated verified tweet

0

1

Updated 2021-05-16

Contributors are: