Information Seeking Behavior and Mental Health Service Utilization: Using Big Data Tools to Examine Temporal Trends and Geographic Variations in ADHD 

Persistently low rates of children’s mental health service utilization have inspired close examination of barriers to care that point to sociodemographic and geographic disparities. Information science points to socioeconomic disparities in health information seeking (access and need) that may decrease corresponding to increasing rates of online searching in underserved communities. Three specific aims were examined: Aim 1. Examine changes in information seeking over time; Aim 2. Examine geographical variations of online searches; Aim 3. Examine the connection between state-level information-seeking variations and individual diagnoses.  

The dissertation uses publicly available data and big data methods (i.e., time series analyses, machine learning approaches, multilevel modeling) to examine and improve the speed and reach of scientific communication. Time series analyses revealed that 1) queries of “ADHD medication” increase, while queries for “ADHD therapy” remain relatively low despite a positive linear trend, 2) breaks coincided with a decrease in search interest, while post-break periods illustrated a rise, and the ADHD Awareness Month (October) coincided with a rise of public interest in all four search terms. Machine learning algorithms suggested that seeking ADHD-related information online was relatively more important in states with a higher percentage of underserved families (e.g., Hispanic/Latinx youth) and/or with more families who are alreadyconnected to systems of care. Multilevel modeling analyses revealed that racial/ethnic disparities in ADHD diagnoses remain and state-level search interest positively predicted ADHD diagnoses after controlling for sociodemographic variables.

The anonymous and accessible nature of seeking information online makes search engines like Google important sources of mental health information, especially among underserved and marginalized groups. Findings suggest need for future research and highlight internet-based opportunities for understanding and detecting inequalities in need for and access to empirically supported information and care. 

Major Professor: Dr. Stacy L. Frazier