Data-logging exercises in science classrooms assume that with the proper scaffolding and provision of contexts by instructors, pupils are able to meaningfully comprehend the experimental variables under investigation. From a case study of knowing and learning in a fish hatchery using real-time computer statistical software, we show that appreciating the distributions of fish weights and lengths were mediated by the practical understanding that comes from measuring and handling the animals over time. This prior knowledge was further mediated by the representations in mathematical forms where technology was central. Hence, as a fish culturist "played" with the data logging-like technology, expanded forms of learning were made possible during the process of analysis and reflection. Besides clarifying the process of learning when using educational media like data loggers, we present the computer as a self-reflexive tool that constitutes--in a dialectical fashion--learning in the everyday world. (Contains 5 figures and 1 note.)