OHuBB Scientific Meetings

Abstract

Scientific progress depends not only on discoveries but also on knowing when attractive hypotheses fail. This talk examines the “file drawer problem” and the consequences of selective publication of positive findings. I will show how small samples, inadequate control conditions, and flexible analytical choices can produce influential claims that later prove difficult to replicate. I will discuss our work on action video game training in dyslexia and on the neural noise hypothesis, illustrating how null results can be made scientifically informative. Finally, I will highlight why publishing null results is essential for cumulative and reproducible science.

About the speaker

Katarzyna Jednoróg is a neuroscientist and head of the Laboratory of Language Neurobiology at the Nencki Institute of Experimental Biology, Polish Academy of Sciences. Her research examines how language and reading skills emerge in childhood, how their neural networks develop, and how and why these trajectories differ in neurodevelopmental disorders. She also studies experience-dependent plasticity across sensory and linguistic contexts, including in blind and deaf individuals. Her work combines behavioral assessment with multimodal neuroimaging, including functional and structural MRI, diffusion and quantitative MRI, magnetic resonance spectroscopy, and EEG, with an emphasis on rigorous, reproducible developmental cognitive neuroscience.