Automated Quality Assessment of Depression Websites
Author Information
Author(s): Kathleen M Griffiths, Thanh Tin Tang, David Hawking, Helen Christensen
Primary Institution: Centre for Mental Health Research, The Australian National University
Hypothesis
Can an automated quality assessment procedure effectively rank depression websites based on their evidence-based quality?
Conclusion
An automated system can differentiate depression websites of varying evidence-based quality, which could significantly aid consumers in finding reliable health information online.
Supporting Evidence
- The correlation between the AQA score and the evidence-based ratings was high and significant (r=0.85, P<.001).
- The AQA explained 82 percent of the variance in evidence-based quality scores.
- Google PageRank was found to be a weaker indicator of evidence-based quality compared to the AQA.
Takeaway
This study created a computer program that helps find good websites about depression, making it easier for people to get reliable health information online.
Methodology
The study developed an Automated Quality Assessment (AQA) procedure and validated it against human ratings of 30 depression websites.
Potential Biases
One of the evidence-based raters was involved in site selection, which could introduce bias.
Limitations
The AQA is focused only on treatment information and may not reflect the quality of other types of information on a site.
Participant Demographics
The study evaluated 30 depression websites, with characteristics including ownership structure and involvement of health professionals.
Statistical Information
P-Value
P<.001
Confidence Interval
95% CI = -0.96–0.62
Statistical Significance
p<0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website