In this paper, we discuss the individuals' "roles", "responsibilities", and "routine activities", along with their "goals" and "intentions" in two different contexts--a school science context and a university research context--using sociological lenses. We highlight the distinct characteristics of both contexts to suggest new design strategies for STEM learning environments in school science context. We collected our research data through participant observations, field notes, group conversations, and interviews. Our findings indicate that school science practices were limited to memorizing and replicating science content knowledge through lectures and laboratory activities. Simple-structured science activities were a means to engage school science students in practical work and relate the theoretical concepts to such work. Their routine activities were to succeed in schooling objectives. In university research settings, the routine activities had interdisciplinary dimensions representing cognitive, social, and material dimensions of scientific practice. Such routine activities were missing in the practices of school science. We found that the differences between school and university research settings were primarily associated with individuals' "goals" and "intentions", which resulted in different social structures. In school settings, more authentic social structure can evolve if teachers trust their students and allow them to share the social and epistemic authorities through establishing mentorship. We do not expect school science students to perform the tasks of scientists in the same manner, yet the desired school science activities should include mentorship roles and interdisciplinary perspectives and encourage school science students to pursue unanswered questions without looking for the right answer.