This paper aims to determine the distribution of problem spaces in learning activities, when geovisual analytics is introduced into social science education. We know that various dimensions of complexity emerge in learning activities including this kind of technology. This paper clarifies the features of the problem spaces in such activities. The study was conducted in three middle schools in Sweden, in four social science classes with students aged 10 to 13 years. The specific geovisual analytics platform used was Statistics eXplorer. The learning activities were followed for two to four weeks at each school using video observations. Drawing on actor–network theory, we conducted material discursive analyses of the learning activities. The geovisual analytics generally support student understandings, but the didactic design of the classroom was not completely supportive. Six central aspects were found in the distribution of problem spaces within the learning activities. Novel approaches to pedagogy and teaching employing geovisual analytics could benefit students’ knowledge building as they work with visualized data.