Searching for specific health‐related information in MedlinePlus: Behavioral patterns and user experience

Abstract

Searches for specific factual health information constitute a significant part of consumer health information requests, but little is known about how users search for such information. This study attempts to fill this gap by observing users' behavior while using MedlinePlus to search for specific health information. Nineteen students participated in the study, and each performed 12 specific tasks. During the search process, they submitted short queries or complete questions, and they examined less than 1 result per search. Participants rarely reformulated queries; when they did, they tended to make a query more specific or more general, or iterate in different ways. Participants also browsed, primarily relying on the alphabetical list and the anatomical classification, to navigate to specific health topics. Participants overall had a positive experience with MedlinePlus, and the experience was significantly correlated with task difficulty and participants' spatial abilities. The results suggest that, to better support specific item search in the health domain, systems could provide a more "natural" interface to encourage users to ask questions; effective conceptual hierarchies could be implemented to help users reformulate queries; and the search results page should be reconceptualized as a place for accessing answers rather than documents. Moreover, multiple schemas should be provided to help users navigate to a health topic. The results also suggest that users' experience with information systems in general and health-related systems in particular should be evaluated in relation to contextual factors, such as task features and individual differences.

Seeking treatment options: Consumers' search behaviors and cognitive activities

“…Our first research question was: what behaviors do consumers exhibit when they search for treatment options online? In many ways, participants' behaviors when searching for treatment options were consistent with consumers' health information search behaviors reported in numerous prior studies (e.g., Spink et al, 2004; Zeng et al, 2006; Zhang, 2013). That is, people submitted short queries, misspelled keywords, and selected results exclusively from the first page of search results.…”

“…The basic level was consumers' behavior of using systems to searching for health information. Common findings about system use from prior studies include: Consumers typed in short queries, with the average query length ranging from one to three terms (Spink et al, 2004); consumers had difficulties in articulating their needs using keywords and misspelled search terms (Zeng et al, 2006); and they examined few search results, particularly when they look for factual information (Zhang, 2013).…”

Toward a layered model of context for health information searching: An analysis of consumer‐generated questions

“…By analyzing a large number of logged search sessions, Cartright, White, and Horvitz () found that if the users' goal was to look for evidence, they were more likely to look for details and relevance of symptoms; if the goal was to test a hypothesis, they were more likely to pursue content on facets of one or more illnesses, including risk factors, treatments, and therapies, and on the differentiation between diseases. Zhang (in press) also found that when the users' goal was looking for specific factual information, they were less likely to modify queries and examined very few search results; they also demonstrated expectations of finding the answer on the search results page. Second, the complexity of search tasks was found to influence search behaviors.…”

Aggregating semantic information nuggets for answering clinical queries

“…It has several peculiarities mainly related to the search process patterns [4] and the structure/vocabulary of the queries [23]. More specifically, expert clinical information searches are generally performed by clinicians within the Evidence-Based Medicine (EBM) approach [16].…”