Along with the popularization of digital technologies in education from the early 1980s, there has been a practice of evaluating the effects of educational technology on educational parameters. A high occurrence of systematic reviews and meta-­‐analyses show that the evidence of the benefits remains elusive, and that there is little scientific evidence to prove the efficiency of new digitally supported learning models. Together these evaluations create insecurity about the contribution of ICT to the learning process. Yet, there is still a strong conviction in political institutions that education can be reformed for the better, by developing strategies that place digital technology at the center of their concerns. The dissertation works with the thesis that producing evidence for the benefits of educational technology is methodologically and philosophically unreasonable. The thesis is examined by analyzing a corpus of white papers and politically commissioned evaluation reports spanning four decades, in order to map arguments, epistemic sources and how arguments are used to recommend political interventions. The corpus consists of publications from influential global institutions like the UNESCO, the European Union, World Economic Forum and the Organization for Economic Co-­‐operation and Development (OECD). Selection of the literature for the literature review and the analysis of the corpus is executed through a realist/anti-­‐ realist prism. This analytical prism seeks to map policy discourse onto a familiar tension between the realist sciences and schools of thought rooted in continental philosophy. The dissertation demonstrates that arguments for the value of ICT in education are dominated by evidence-­‐based methods, but that interpretation of quantitative reports, to a large degree is based on an idealism that is inspired from continental pedagogical theory. This idealism, in turn, affects what is regarded as valid data -­‐ and gradually develops into historically new forms of evidential practice. Hence the title of the dissertation: "Evidence of a potential".