Abstract: PO1073
The Information Dilemma
Session Information
- Educational Research
November 04, 2021 | Location: On-Demand, Virtual Only
Abstract Time: 10:00 AM - 12:00 PM
Category: Educational Research
- 800 Educational Research
Author
- Desai, Tejas, NOD Analytics, Harrisburg, North Carolina, United States
Background
The Internet has changed search. The amount of searchable information is growing as more bytes are added/modified than deleted and the criteria by which search results are obtained has also changed. Results are not exclusively obtained by their relevance to the query. Social media search uses predictive algorithms to display results with which the learner is most likely to engage (retweets, likes, replies, clicks). Results that promote engagement are valued more than those pertinent to the query. Known as customized search, this strategy is obvious when 2 individuals make an identical query and receive different search results and in a different order. Customized search protocols increase engagement but at the cost of creating an information dilemma. In this dilemma, each learner is exposed to a different set of facts upon which scientific discussions are started. In order to establish a common set of facts, I created a search engine based on a standard search protocol.
Methods
NephTwitterArchive.com is a non-commercial search engine that identifies scientific tweets from various NephTwitter communities. I coded 51 algorithms to identify scientific tweets at 15-minute intervals. Scientific tweets have informative text and a slide, URL to a scientific resource, or both. The engine uses a standard search protocol in which search results are based only on the query. All searches are anonymous. I measured learner-affinity for the engine by number of completed searches. Four elements are required for a single completed search to be recorded: query is made & executed, results are displayed, and a link to the primary scientific tweet is clicked.
Results
From 11/2011 to 4/2021, the engine identified 341742 tweets from 517 NephTwitter communities. A third of these were scientific tweets. From 10/2019-4/2021 learners completed 28313 searches (monthly median 1368, IQR 1264-1580). From 2019 to 2020, median monthly completed searches changed +12%; from 2020 to 4/2021, a +38% change. Nearly half of all learners used the engine immediately after visiting Twitter or Facebook; 37% visited the engine directly.
Conclusion
Scientific discourse is valuable if all participants start the conversation with a common set of facts. Social media search tools do not support this goal. A search engine that uses standard search protocol can mitigate the information dilemma and restore search to its primary function of providing results based on query alone.