This dissertation aims to explore the effect of dialogic argumentation on grade rn8 students' epistemic knowledge of science in physics learning and to identify the practices rnand challenges faced to promote epistemic knowledge through dialogic argumentation. The rnstudy used a mixed methods experimental design research approach where a quasi experimental design was employed to compare experimental, and control groups' epistemic rnknowledge of science and a collective case study design was employed to identify teachers' rnchallenges in promoting students' epistemic knowledge in argumentation lessons. Fourteen rnclassrooms were randomly selected from twelve schools and assigned to intervention (239 rnstudents from seven classrooms) and control (240 students from seven classrooms) groups. A rnpre-intervention physics reasoning test was administered to both groups and small group rnclassroom discussions were video recorded. Then, physics teachers in the intervention group rnhad trained for three days about dialogic argumentation and its implementation. In addition, rnTalking Physics Students Activities manual, that contains fifty-two argumenattives physics rnactivities were distributed and used in a yearlong dialogic argumentation intervention. Post intervention data were obtained from a physics reasoning test administered to both groups, rnvideo records of small group classroom discussions from both groups and whole-class rnteaching and audio records of teacher interviews from intervention group. The test scores and rnthe quantized qualitative data of small group discussions were analyzed using inferential rnstatistics. Video data of classroom teaching were quantitized using quantitative ethnography rnand analyzed using epistemic network analysis. Teachers interviews were also qualitatively rnanalyzed using a thematic analysis. Mann-Whitney U test results indicated that the post-test ivrnscores of grade 8 students in the argumentaion lessons significantly increased in their level of rnepistemic knowledge compared to the non-argumentation groups, z = - 4.509, p = .000, and r rn= .21, but not in the pre-test scores, z = - 1.038 and p = .299. However, both pre- and post test scores of both groups were relatively low. The intervention groups showed significant rnimprovements in the quality of their argumentation on the ASAC scale, z = 2.111, p = .035, rnand r = .56, but not the control groups, z = 1.068 and p = .285. The epistemic network rnanalysis of the wholeclass teaching in the intervention group showed weak and less frequent rnconnections epistemic aims, epistemic processes of construction, justification, and evaluation rnof knowledge. The study found evidence that argumentation-based lessons improved both the rnepistemic knowledge and the quality of dialogic argumentations of grade 8 students and that rnstudents' level of epistemic knowledge and the quality of their dialogic argumentations were rnstrongly correlated. However, teachers had failed to teach argumentation as an epistemic rnpractice. To make a better use of dialogic argumentation, therefore, well-thought-out and rnresearch-supported training needs to be given to science teachers as part of their continuous rnprofessional development. Future studies are necessary to address the effects of such rntrainings and to determine if there are other hidden factors, apart from teachers' inadequacy, rnthat affect students' epistemic knowledge of science.