Understanding the preference of online health information seeking among college students using the best-worst scaling method

Dan Wang, Jiang Wang, Haihong Chen, Manli Chen, Guangwen Gong, Yining Sun, Yonghui Wu, Xuemei Wang, Xiping LiFrontiers in Public Health 2025 · DOI

TL;DR

College students heavily rely on online health information, but misinformation is rampant. This study used a novel "best-worst scaling" method to pinpoint what factors are most crucial for students when seeking health information online. The findings highlight a strong preference for verified, trustworthy sources and guaranteed privacy, suggesting targeted interventions are needed to improve the quality and reliability of online health information for this demographic.

What Problem Does This Paper Address?

The internet has become a primary source for health information, offering unprecedented access to knowledge about diseases, treatments, and lifestyle choices. However, this accessibility comes with a significant challenge: the overwhelming prevalence of health misinformation. This is particularly concerning for college students, who are in a critical developmental stage for making informed health decisions and are highly digitally immersed. While previous research has identified various factors influencing online health information seeking, there's a lack of clear understanding regarding the relative importance of these factors for college students. Simply listing factors isn't enough; we need to know which ones truly matter most to guide effective interventions.

Methods

The researchers employed the Best-Worst Scaling (BWS) method, a powerful technique for determining preferences and relative importance. Unlike traditional rating scales that can suffer from response biases, BWS forces participants to make trade-offs by repeatedly selecting the "best" and "worst" options from a set of attributes.

Here's how it worked:

  1. Attribute Identification: An initial list of potential factors influencing online health information seeking was compiled through a comprehensive literature review and a focus group with college students.
  2. Attribute Refinement: This list was refined to 11 key attributes, which were then described in detail (e.g., "Information source from trustworthy and authoritative website," "Verified by professional institutions or health professionals," "Privacy and security guaranteed").
  3. BWS Questionnaire Design: A balanced incomplete block design (BIBD) was used to create 11 distinct tasks. Each task presented participants with a subset of 5 attributes, and they were asked to choose the single most important and the single least important attribute from that subset.
  4. Data Analysis: The collected BWS data was analyzed using both a counting approach (best-minus-worst scores) and a mixed logit model. The mixed logit model was particularly important as it accounts for heterogeneity in preferences among respondents. The study also explored the influence of having a medical education background on these preferences.

Datasets

The study utilized data collected from an online survey administered to college students from four universities (two medical, two non-medical) in Hubei and Jiangsu provinces, China, between March and June 2023. A total of 733 undergraduate students were invited, and 534 provided valid responses after exclusions. No new benchmark datasets were introduced; the study relied on existing literature for attribute generation and standard BWS analysis techniques.

Key Insights and Ideas

  • Trust is Paramount: College students overwhelmingly prioritize the credibility and trustworthiness of information sources. Attributes like "Verified by professional institutions or health professionals" and "Information source from trustworthy and authoritative website" were ranked as the most important, collectively accounting for a significant portion of their preferences. This suggests a strong desire for authoritative validation in a sea of online information.
  • Privacy as a Core Concern: Beyond just accuracy, privacy and security emerged as a highly valued factor. This indicates that students are aware of and concerned about how their personal information is handled when seeking health information online.
  • "Free" Isn't Always Best: The attribute "without advertisement links" was consistently ranked as the least important factor. This implies that students are willing to tolerate advertisements if the information itself is deemed reliable and trustworthy, prioritizing substance over an ad-free experience.
  • Medical Education Nuances: While medical students didn't necessarily prioritize accuracy and credibility more than their non-medical peers (suggesting they already possess these skills), they showed a greater preference for "writing and language" and "professional interface design." This hints that for those with a medical background, the presentation and clarity of information become more critical.

Key Results

  • The top four most important factors influencing online health information seeking among college students were:
  • "Verified by professional institutions or health professionals" (mean BW = 1.938)
    1. "Information source from trustworthy and authoritative website" (mean BW = 1.921)
    2. "Privacy and security guaranteed" (mean BW = 1.234)
    3. "Consistency of information" (mean BW = 0.803)
  • The least important attribute was "without advertisement links" (mean BW = -2.618).
  • Mixed logit model results confirmed these rankings, with coefficients showing significant positive preferences for the top attributes.
  • Receiving medical education had a positive effect on preferences for "writing and language" (coefficient = 0.410) and "professional interface design" (coefficient = 0.279), but a negative effect on "disclosure of author information" (coefficient = -0.307).

Limitations and Future Work

The authors acknowledge several limitations:

  • Attribute Scope: The study's findings are limited to the 11 attributes included in the BWS questionnaire. Other potentially important factors might not have been captured.
  • Selection Bias: While efforts were made to recruit a diverse sample, selection bias is always a concern in online surveys.
  • Generalizability: The study was conducted in China, and while the findings align with some international trends, further research with more representative samples across different cultural contexts is recommended.

Future work could explore the impact of social media platforms more deeply, investigate the role of digital health literacy interventions, and expand the study to a broader range of age groups and populations. The authors also suggest that interventions should focus on promoting accuracy, credibility, privacy, and consistency in online health information management for college students.

Summary updated

Abstract

We recommend that concerned authorities consider interventions targeting the accuracy, credibility, privacy, and consistency of online health information management for college students.
BibTeX
@inproceedings{wang2025understanding,
  author = {Dan Wang and Jiang Wang and Haihong Chen and Manli Chen and Guangwen Gong and Yining Sun and Yonghui Wu and Xuemei Wang and Xiping Li},
  title = {{Understanding the preference of online health information seeking among college students using the best-worst scaling method}},
  booktitle = {Frontiers in Public Health},
  year = {2025},
  doi = {10.3389/fpubh.2025.1670106},
}
Figure