Twitter Posts, Gist, and the Perceived Risk of Using a Novel Recreational Drug

Theoretical Background and research questions/hypothesis:  Individuals frequently encounter two types of information when evaluating the risk of using new recreational drugs: base-rate evidence describing the frequency of positive and negative outcomes, and anecdotal evidence describing positive and negative personal experiences. The present study evaluated the relative importance of both types of evidence when considered simultaneously. Participants created a twitter post with a maximum of 140 characters describing a novel recreational drug to their friends. These twitter posts reflected the key information (i.e., gist) that participants extracted from the two types of evidence (base-rate and anecdotal). It was predicted that negative anecdotes would minimize the impact of base-rate evidence addressing the potential safety of the drug.

Methods:  Two hundred and ten adults were randomly assigned to one of four conditions. Participants in each condition were told that a study of 1,000 college students revealed that 95% had experienced positive effects from the drug. Each condition varied the number of positive and negative anecdotal reports that students read about the drug. These anecdotal reports were selected from web postings archived on Erowid.org, an online platform that reports personal drug experiences as well as drug related research findings. Participants in Condition I only received base-rate information regarding the drug’s effects. Participants in Condition II received the same base-rate information as well as five negative anecdotal reports and one positive anecdotal report regarding the drug. Participants in Condition III received base-rate information as well as three positive and three negative anecdotal reports. Participants in Condition IV received base-rate information as well as five positive anecdotal reports and one negative anecdotal report.  After reading the assigned material, participants created a Twitter post (140 characters or less) that would allegedly be sent to their friends telling them about the drug. Participants also completed items assessing the perceived harmfulness of using the drug, the perceived enjoyment of using the drug, and the perceived likelihood of a favorable or adverse outcome when using the drug.

Results:  Experimental condition significantly influenced the perceived likelihood of experiencing a favorable outcome when using the drug, F(3, 202) = 12.39, p < 0.01,  perceived likelihood of experiencing an adverse outcome when using the drug, F(3, 202) = 16.787, p < 0.01, the perceived harmfulness of using the drug, F(3, 201) = 5.23, p < 0.01, and the perceived enjoyment of using the drug F(3, 201) = 8.175, p < 0.01. The impact of experimental condition on the valence (positive and negative) of the twitter posts produced by participants will also be reported. 

Conclusions:  The current findings suggest that young adults overweight anecdotal evidence regarding the risk of a recreational drug even when presented with conflicting base-rate evidence. 

Implications for research and/or practice:  Findings from the current study have implications for designing health risk communications regarding the use of emerging recreational drugs. The findings also have implications for using twitter posts to identify the kernel of meaning (gist) that young people extract from health related risk messages.