Abstract
Four experiments with 4th, 5th, and 6th graders addressed conceptual change in response to anomalous data about empirical regularities in science. Impedance to conceptual change in response to anomalous data could potentially occur at any of 4 cognitive processes: observation, interpretation, generalization, or retention. In the 4 experiments, conceptual change was blocked most strongly at observation. The students had difficulty making accurate observations, but they did not simply observe what they expected to observe. The students usually aligned their conceptions with their observations, evincing an implicit epistemology in which they distinguished conceptions from evidence and changed beliefs in response to evidence. Providing explanations to students promoted conceptual change, evidently because the explanations provided a schema that guided the observation process.