There is substantial research in science education about students', teachers', and scientists' views of nature of science (NOS). Many studies have used NOS frameworks that focus on particular ideas such as tentativeness of scientific knowledge and cultural embeddedness of science. In this paper, we investigate NOS from the perspective of the Family Resemblance Approach (FRA) which considers clusters of ideas about science in terms of categories that offer a comprehensive analytical lens to studying NOS views. The empirical study re-analyzes NOS views obtained from 7 and 8th grade students, science teachers, and scientists using the FRA lens. Statements from all three groups were obtained using a free-write questionnaire on nature of knowledge and nature of knowing. The statements were reclassified using the FRA framework. Epistemic network analysis (ENA) was applied to the statements produced by each group of participants, and the resulting network models were interpreted and compared. The results show that student and teacher network models possessed no central idea, and more tangible ideas about science were frequently connected. Scientist network models showed more connections across their statements which indicate a higher degree of agreement and coherence among a variety of ideas compared to student and teacher network models. The paper discusses the findings as well as the methodological contributions, and concludes with implications for future research.