<![CDATA[Linguistically, the more complex clause a sentence implements, the more complex the grammar it embed. A simple clause, as opposing the complex clause, consists of a subject and a verb with optional object or complement. However, Biber et.al. (2011) has proved that complexity of a clause attached in a sentence fails to indicate the complexity of a sentence in writing composition. Furthermore, they indicated the stage of the acquisition of complexity features of writing, including that of in L2 of English. This present study aims at investigating how the frequency of complex sentence produced by Ganesha University of Education students different from that of by professional academic writer, to assess studentsâ writing ability. The complexity features of writing comprise: 1) fnite dependent clause types; 2) non-fnite dependent clause; and 3) dependent phrase types (non-clausal). Theoretically, the implication of the study aims to beneft the development of teaching writing through authentic assessment process. Providing assessment with diagnostic feedback on areas of needed improvement through authentic assessment in English language teaching has become a solution due to its signifcance for language learner (Brown & Abeyvikrama, 2010: 254). Corpus-based analysis is conducted allowing great number of data to be analyzed to generate more reliable generalization (Baker, 2010). The corpus data and the frequency of sentence complexity as represented by selected features of (Biber et al., 2011), are annotated using CLAWS part-of-Made Wahyu Mahendra, An Authentic Assessment....46. speech tagger hosted by Ucrel, and are calculated by Ant.Conc 3.2.4 corpus software. The fndings projects surprisingly occurrence in which undergraduate students utilizes clausal sentence rather than phrasal sentences. It becomes indication that the students are in the early developmental stage of complexity in writing. ]]>
