Over the past several decades, the TESOL community in general has seen an increase in the interest in the use of data-driven learning approaches. Most of these have focused on the acquisition of vocabulary items, including a wide range of information necessary for their correct usage. One type of vocabulary that has been investigated has been that used mostly in academic writing. However, there has been little empirical study into the acquisition and use of two types of vocabulary common in academic writing, linking adverbials and reporting verbs. In addition, most studies into data-driven learning and academic vocabulary acquisition have assumed that the acquisition of this vocabulary will automatically lead to better academic writing. However, there have been no studies which have confirmed these assumptions. This study will attempt to fill these gaps in the literature. It investigates the acquisition of linking adverbials and reporting verbs, two sets of vocabulary that are difficult for ESL writers to master, with the use of data-driven learning techniques. In addition to looking at the acquisition of this vocabulary specifically, this study investigates whether these approaches also contribute to students' progress in their academic writing. Results indicate that data-driven learning techniques do in fact help students acquire and better employ linking adverbials in their writing, although they do not help in the case of reporting verbs. No correlations to increases in student essay scores were found. However, this study still shows the possible value of data-driven learning approaches to students' academic writing.