SQL is a typical query language for performing data analytics. Although its usage is ubiquitous, learners ex- perience that query formulation in SQL is error-prone and time-consuming. Prior research has shown that this is due to low expressive ease, extensive training requirements and high cognitive load, all of which present a significant burden for SQL learners. Visual representations can assist learners to significantly lower this burden. The current dominant paradigm aims to facilitate SQL querying by helping users to avoid the syntax of SQL. Such Visual Querying Systems (VQS), however, are not effective for SQL learners as they hide the syntax of the language during query formulation, rather than assisting learners to write correct queries in SQL. Furthermore, training with VQSs is system specific, which leads to system dependency for learners. We argue that novices need support from Visual Query Representation (VQR) solutions which, instead, help them in learning how to write correct and portable SQL queries. In this paper we present SQLVis, a VQR to support novice SQL users in query writing. Our system represents the query as written in SQL by the user, which can improve the SQL writing proficiency of its users. Results of an in depth empirical study demonstrate the significant value of SQLVis for learners.