The first intellectual output of our project was initially defined as the Mainframe Database. However, during development, the project partners decided to give it a more user-friendly title so we renamed it as our Task Builder. The back-end of this output remains  a relational database containing all language tasks (or questions) that will be delivered in a placement test. The font-end, however, provides an intuitive and interactive question builder, incorporating  a number of different question types (i.e. multiple choice, gap-fill, drag-and-drop, etc.). This allows users to create their own questions for each of the six languages supported by the NGPT app at this time.

The output was actually upgraded quite a bit from the original  design. First of all, by introducing multiple question types for the Use of Language questions, the partners realised that one particular type could, in fact, provide different problem-solving techniques. For example, a drag-and-drop question could have a single moving part, allowing us to create questions where the test-taker must drag specific word into the correct part of the sentence. But if we mark all parts as movable, we can have questions where test takers must form a sentence using the  entire list of words provided. As a result of these changes, we also needed to implement a system of “task instructions”, which provide guidance for test takers.

Next, we decided that in addition to Use of Language questions, the placement test should contain both reading comprehension and listening questions. While this significantly increased the scope of the development work, the partners decided they should implement these additional tools into the project.

Finally, our developers implemented a sophisticated multi-tenant user management system. This allows institutions to create an umbrella user account and subsequently create individual user accounts for their teachers. This way we can track database entries at an  institutional level.

All features were reviewed in extensive quality control and testing and their development was fully completed by the end of 2021. In addition, the partners jointly created almost 10.000 unique Use of Language questions  for the six target  languages. In the upcoming period, we will be adding the reading comprehension and listening questions to the database, while our developers work on adjusting the ELO ranking system for our needs.