A SYSTEMATIC REVIEW OF METADISCOURSE MARKERS IMPLEMENTATION IN WRITTEN COMMUNICATION
The main objective of conducting this systematic review is to study the usage of metadiscourse markers in expository writing among ESL and EFL novice writers by reviewing pertinent aspects such as the types of metadiscourse markers used, the effect of language proficiency on metadiscourse usage and the types of language errors committed from the perspective of metadiscourse studies. To achieve the objective, the literature published in four different databases was reviewed. After applying a series of selection criteria, 26 empirical articles were selected including a total of 1696 novice writers in English writing with ages ranging from 15 to 59 years and the analysis of 17,777 essays. These students were evaluated with two categories of metadiscourse markers namely interactive and interactional as well as taking their language proficiency and language errors committed into consideration. The findings from this review show that (1) there are 86% (19 out of 22 articles) metadiscourse studies conducted on interactional markers as compared to interactive markers; (2) 43% (6 out of 14 articles) metadiscourse studies done on high proficiency students as compared to low and average proficiency students, and (3) 66% of the language errors commonly associated with undergraduate students are the misuse of metadicourse markers (6 out of 9 articles). This systematic review would benefit future researchers in determining the research direction based on the gaps discovered from the previous study and language educators in catering their learners with different proficiencies of the proper use of metadiscourse markers. Additionally, this systematic review would give priority to aspects of metadiscourse that have received less attention, such as interactive metadiscourse markers, the use of metadiscourse markers by learners who are less proficient, and the necessity of a framework that focuses on the mistakes made when using metadiscourse markers rather than language in general.