This study aims at finding out and comparing students' ability in identifying abuse(s) in argumentdeductions between science students and social science students. The design of this study is a descriptivequalitative. There were 108 students involved as the respondents (52 science students and 56 socialscience students). The instrument utilized was ten arguments taken from Guth (1969), which is aspecified test to drill argument analysis—aligning content validity for this current study. Therespondents were asked to analyze these arguments and find out the deduction abuses. Their approachin analyzing each argument was further interpreted through data analysis. There were a total of 1080analyses, but 477 analyses of which were discarded due to a biased approach. The data were analyzedusing thematic and interactive analysis. The result shows that, among science students, the mostemployed approach is faulty premise (199 analyses), followed by misleading statistics (53 analyses),hidden premise (37 analyses), equivocation (10), and circular premise (4 analyses). Meanwhile, amongsocial science students, the majority also exploited the faulty premise (137 analyses). Additionally, thehidden premise was also engaged in a great number (130 analyses), equivocation (40), followed bymisleading statistics (9 analyses), and circular premise (2 analyses). These findings circumstantiallyimply that, in learning, students with a science background are better at capturing stated details, whilestudents with a social science background are competent at spotting both stated and unstated details inarguments. It is suggested that teachers should balance the students’ reasoning approaches, regardlessof their academic backgrounds to achieve learning objectives.Keywords: logic, students’ cognition, deduction, reasoning skills, and teaching and learning.