Evaluation of a novel Conjunctive Exploratory Navigation Interface for consumer health information: a crowdsourced...

Area of Science:

• Health Informatics
• Human-Computer Interaction
• Information Retrieval

Background:

• Consumer health information websites are abundant but often underutilized due to insufficient search functionalities.
• Lookup search alone is inadequate for consumers to fully leverage available health information resources.
• Exploratory search presents a promising alternative for enhancing consumer health information retrieval.

Purpose of the Study:

• Introduce a novel Conjunctive Exploratory Navigation Interface (CENI) designed to improve consumer health information retrieval and navigation.
• Evaluate the effectiveness of CENI by comparing it against traditional search interfaces using crowdsourcing.

Main Methods:

• Collected over 60,000 consumer health questions from NetWellness.
• Developed CENI, a topic-navigation-based interface, and a keyword-based lookup interface for comparison.
• Conducted a crowdsourced comparative study with 90 Amazon Mechanical Turk workers evaluating three search modes (CENI, keyword lookup, Google/NetWellness advanced search) across 9 tasks.

Main Results:

• The Conjunctive Exploratory Navigation Interface (CENI) was found to be significantly easier to use than both keyword-based lookup and standard search interfaces (P<.001).
• A majority of participants (47.8%) preferred CENI, citing its ease of use, organized topic-based content, and ability to quickly narrow down information.
• CENI effectively supported exploratory navigation for users unsure how to begin their search.

Conclusions:

• A novel Conjunctive Exploratory Navigation Interface (CENI) was developed for consumer health information retrieval.
• Crowdsourcing provided a timely and cost-effective method for evaluating search interfaces with a large participant pool.
• This study demonstrates that combining exploratory navigation with lookup search is a more effective strategy for consumer health information retrieval than lookup search alone.