What do researchers cite when reviewing past research? a semantic analysis of citation use in literature reviews of research articles in information systems

Abstract

Citation is an essential feature of research writing, which is particularly expected in the Literature Review (LR) to situate the writer's work among others' in the field. Yet, making references to past scholarship in the genre-part can present different sorts of challenges to novice writers with the major one being having to choose not only appropriate source ideas but also the right amounts to cite. It is not uncommon to find indiscriminate citing criticized for being bibliomaniac (Belcher, 1994), losing authorial control (Ridley, 2012) or laundry listing. Although much research has been conducted on citation practices and numerous guidebooks about the topic have been published, most of these sources have focused on citing mechanics (e.g., citation forms), citation verbs, and generic citation functions rather than what needs to be cited and how cited ideas function in specific parts of a research text such as its LR. Addressing this gap, this paper will present a study of source ideas cited in LRs of 30 research articles in Information Systems that follow a positivist paradigm. Adopting a move/step-specific approach to the analysis informed by Swales’ (1990) CARS model and the move-specific semantic-functional citation typology developed by Kwan and Chan (2014), the study examined the source ideas in terms of their semantic attributes and their use in specific moves and steps of the LRs. Findings show distinct patterns of ideas referred to in the steps. Given the space, this paper will only discuss those observed in Move 1 Step 3 (Reviewing items of previous research), which, as observed, mainly refer to four types of ideas with distinct semantic attributes. The findings can be attributed to the rhetorical purpose of the step, what the writers try to accentuate in Move 2, and equally importantly, the paradigm assumed in the studies. Pedagogical implications will be discussed.