At my alma mater, our Computer Science department requires all graduating students to take an assessment test after their first two years and a second test as they prepare to graduate. The tests were created internally and consists of questions culled from exams used by various department faculty. It costs the student nothing out of pocket except for registration and testing time.

I should mention at this point that when we instituted these assessments, it took us several semesters to decide that we needed to establish a minimum passing score. A handful of students were simply choosing A as the answer for every question and completing a 100-question assessment in 5 minutes. We needed to help motivate our students to do their best, so we established a minimum passing score of 60%.

Most of our students just take the test as a rite of passage and get it over with. However, a few of them objected, stating that a multiple choice test about details of a certain programming language does not demonstrate programming proficiency. Of course, if the test were easy, there would be few complaints and you and I wouldn’t be having this pleasant conversation. A small minority of our students score less than 50% on the exam and have to retake it. It is this minority that are the vocal complainers.

Study “materials” are available, consisting mostly of a list of topics that might appear on the exam. The exam questions themselves come from a pool of questions and if you retake the exam, you will likely receive a different set of questions from the pool. Students failing the exam also claim the study materials are too vague to help them prepare.

All of this is standard procedure for testing at a college level, but what exactly does passing this test prove? I think it proves that you are a good test taker. Neither of these assessment tests involve the coding and testing of a complete application. Some of our students are transfer students who took their first two years of Computer Science at another institution whose credits we accept. Our tests contains many detailed questions about programming languages these students never encountered in their prior work. Students who don’t do well on our assessments are frequently transfer students who claim the test is biased.

The vast majority of our students graduate from our program well prepared to begin a career in software development, as reported by employers. Although they might encounter some sort of entrance exam given by the interviewing team, typically these exams test the applicant’s ability to solve problems and to validate what they claimed on their resume.

Now, we are an accredited university with an accredited Computer Science program. This means that assessment is a fundamental component of our curriculum. But how are we helping our students succeed in the job market and in their careers with such assessment tests? Our tests are not standardized, but neither is our curriculum. Our department prefers to remain flexible and agile to keep up with trends in the industry with input from our Industry Advisory Council.

And here is the really tough question: should we deny a diploma to any student that does not pass these assessment tests?