During the twentieth century, an increasing amount of data was collected in connection with engineering of the modern states including their education systems. In the Nordic welfare states post WW2, such data were also generated in experimental projects aiming at reforming curriculum and education, and the data were accumulated either as documentation of the pedagogical experiments that took place or as a knowledge base paving the way for designing pedagogical experiments, or both. The article addresses the methodological challenges when analysing the archived data collected as part of experimental projects concerning children's expressions of their "life questions" in Sweden following the big school reforms in the 1960s and continued in altogether six projects conducted from the late 1960s and up until the late 1990s. Based on the case of the UMRe (1969-1973), the first archived experimental project that was closely related to the 1960s education politics and education reforms, the article sheds light on the following questions: how should one understand such classroom data historically? How should one understand the archived data in their political and institutional contexts? Whose voices are possibly speaking in the data about "children's life questions?"