Student Report Cards - Redefined



The advent of natural language processing and natural language generation services within the education sector is set to address a number of everyday problems and challenges that are faced by teachers, support teams and administrators in schools, colleges and universities. In this short article I would like to examine how these services will ease the production and distribution of the online student report card.

Identifying the problem

Online student report cards are typically produced in the following manner. Teachers either type comments or they select from a set of predefined statements about each student's progress. A graphical or tabular dashboard will also accompany this text. So what are the issues surrounding the production of the student report card?

  • the production of student report cards for a single class can take many hours. The time taken to produce them is compounded when student reports cards have to be produced for the entire school or college. There is always the issue that teachers may not produce the reports on time;
  • the reports may contain errors or omissions. This typically increases as the teacher starts to work on the 50th, 60th or 80th student report card;
  • the frequency of producing student report cards is typically set to one per term. Whilst many parents welcome the termly report card others may wish to have a more frequent insight into the progress that their son or daughter is making with their studies; and
  • student facing or parent facing dashboards have been common place for many years. They typically display graphs, charts or tables on attendance, coursework marks and exams grades. Additional commentary by teachers is either entered or selected manually by teachers. As before, this could lead to spelling mistakes, grammatical errors, factual errors or omissions.

The solution

Student or parent facing dashboards already provide part of the solution. The graphs, charts and tables that are presented to the viewer enables them to see up-to-date statistical information regarding the progress of a given student. The dashboard typically presents information relating to a student's attendance, punctuality to classes, coursework grades and exam results. The production of these dashboards is dependent on teachers and support teams keeping up-to-date electronic records of each student in their care.

Parents welcome the insight that this information gives them on the progress that their sons or daughters are making at school or college. However, they are particularly interested in the commentary that is written by the teachers and support teams who support their children with their studies. This commentary is important because it provides parents with answers to the following questions:

  • is my child enjoying his or her studies?
  • is my child on schedule to complete his or studies?
  • how well is my child doing in comparison to his or her peers?
  • what part of the study programme is my child doing well in or not doing so well in?
  • what interventions or support measures have been put in place to support my child at school or college and have the outcomes been positive?
  • is my child taking part in other activities on the campus such as sport, editing the school's online newspaper or volunteering?

The compilation of this narrative is complicated by all the problems that were identified earlier in this article.

The use of natural language generation provides schools and colleges with an opportunity to ease the operational issues that are encountered when producing the student record card. If the technology is used to its fullest extent, the production of the student record card can be completely automated.

student report card

How does natural language generation work in an education setting?

In its simplest form, natural language generation converts data into text by following a set of rules that have been defined by teachers, support teams and managers in a school or college. Here are a couple of examples where natural language generation is used to support the production of a student report card:

Example 1:

Dataset: Student Name: Yasmin. Term 1 Attendance: 86%. Term 2 Attendance: 92%. Intervention: Attendance monitoring team contacted parents. Intervention successful.
Natural language generated text: We are pleased to see that Yasmin's overall attendance has improved from the previous term. Thank you for your support.

Example 2:

Dataset: Student Name: Yasmin. Term 1 Average Grade: Merit. Term 2 Average Grade: Merit. Intended destination: University. Grade required to enter university: Distinction. Events: University Open Day on 5th June 2017.
Natural language generated text: We are delighted to hear that Yasmin wants to go to university. Please note that she needs to raise her average grade from a merit to distinction if she is to secure a place at her chosen university. We are organising another University Open Day on the 5th of June. She may want to keep her options open by looking at other universities who are asking for a merit grade as their entrance qualification.

In both cases the text has been automatically generated using natural language generation services. Teachers, support teams and managers compile the set of rules or algorithms that they wish to use in the generation of the student report card and apply them to the student dataset. When operational the school or college can produce thousands of unique student report cards in two or three seconds.

How will the student report card manifest itself to students, parents, teachers and support teams?

Digital assistants for students like Ada will enable students to review their reports on demand. Since the student report card will update itself as the student dataset changes, students will be able to ask their digital assistant for daily updates. The student report card will also include calls-to-action. For instance, the report card could prompt the student to submit a piece of coursework, to aim for a particular grade for a forthcoming assignment, to book an appointment with the tutor or the careers officer, to apply for student finances, to apply for university, to return a library book and many more.

For teachers and support teams the daily student report card will be automatically generated. Colleagues will receive reports or calls-to-action that classify students into distinct groups; such as those who are at risk of dropping out of their studies, those requiring more support and mentoring, those who are seeking to progress onto further studies or employment and so on and so forth. The use of learning analytics and machine learning agents will also enable teachers and support teams to receive suggestions about what to do to support individual students or groups of students. This advice will be based on the catalogue of historical interventions or actions that have either worked or not worked to support the student body.

For parents, the digital assistant will be the main channel by which to access information about their son or daughter's progress with their studies. They can prompt the digital assistant with any number of questions. They can also use the digital assistant to act as a communication channel between themselves and the staff at the school or college. Parents will continue to access the school or college parental portal for a complete progress report on their son or daughter. The only difference will be that the student report card would have been automatically produced using natural language generation.

For managers, the use of natural language generation will enable better insight into the performance of one class with another or with one department with another. They will receive information that will enable them make better decisions about the resources they will need to support students in coming days, weeks or months. Natural language generation will also mean that managers are no longer spending hours or days compiling, analysing and reporting on the performance of their departments. Instead, more time and energy can be used to address the support needs of individual students in their care.

If schools and colleges can combine the use of natural language processing, natural language generation, machine learning agents and digital assistants they will be in a position to redefine the shape and form of the traditional student report card. The student report card is no longer a static report; but one that updates itself continuously during the course of an academic year. As a result, teachers, support teams, administrators and parents will be in a better position to support students as they progress with their studies. And finally, individual students will be better placed to manage their studies.