Most people today do not read a single article and move on. They scroll through feeds, follow links, watch videos, and absorb information from dozens of sources before forming an opinion. A new paper published in Educational Psychologist argues that researchers studying how people make sense of multiple texts have not kept pace with this reality, and calls for a significant rethink of how online reading comprehension is studied.
The paper, led by researchers at the University of Florida, Lehigh University, the University of Utah, and the University of Texas at Austin, focuses on what academics call multiple text comprehension, the process by which readers draw meaning from more than one source at a time. For decades, studies in this area have placed participants in controlled laboratory settings and handed them a small selection of texts. The authors argue this approach misses much of what actually happens when people read online.
One of the central concerns raised is the role of algorithms. When someone searches for information on Google or scrolls through social media, the content they see is not neutral. Platforms are designed to keep users engaged, which means content that provokes emotion, whether outrage, excitement, or anxiety, tends to spread further and faster than measured, accurate reporting. Readers who are unaware of how these systems work are less equipped to evaluate what they encounter.
The researchers also highlight the importance of culture in shaping how people interpret what they read. Cultural background influences which sources a person trusts, how they approach a task, and what they already believe before they read a single word. Studies that treat prior knowledge purely as a variable to be controlled for, rather than as something deeply embedded in lived experience, risk producing findings that do not translate well to diverse, real-world contexts.
Emotions and affect receive particular attention in the paper. The authors draw a distinction between the narrow understanding of emotion used in most comprehension research, which tends to focus on factors like interest and motivation, and a broader conception drawn from affect studies. In this wider view, readers are often driven by pre-conscious responses to stimulation, impulses that influence which texts they select and how they interpret them, before any deliberate reasoning takes place.
The rise of artificial intelligence adds further complexity. Large language models increasingly summarise information before users ever reach the original source. The authors warn that this reduces exposure to the kind of productive contradiction across texts that helps people build genuine knowledge and critical thinking skills. When summaries smooth over disagreements, readers lose the opportunity to reason across competing viewpoints.
The paper does not dismiss existing research but argues it needs to expand into messier, more authentic territory. Recommended approaches include studying readers in real digital environments, using text sets that reflect cultural diversity, and measuring emotional responses alongside cognitive ones.
Understanding how algorithms, cultural identity, and emotion shape online reading comprehension could help educators design better digital literacy programmes.