Progress testing was developed simultaneously at both sides of the Atlantic Ocean in the 1970s. We will outline how the concept came to birth as a tool aimed to solve the problem of discongruency between a new instructional method (problem based learning) and traditional assessment methods.
A progress test is aimed at the end objectives of an educational program, and repeatedly measures the students’ knowledge with respect to the complete domain of interest of the program. Thus, it enables to monitor the growth of knowledge of individual students and the average growth of cohorts of students across time. Individual growth patterns indicate whether a student is performing well or needs to improve, while average growth patterns provide information on the performance of the educational program. We will demonstrate how this information is provided and how it can be used for benchmarking and to improve education.
Certain requirements should be fulfilled for a progress test to be viable, the corresponding organizational and logistic requirements may be demanding, (lack of) item relevance is a returning issue, the use of a ‘don’t know’ option and negative marking may have its drawbacks, and the efficiency of the measurement may be questioned. In addition to the achievements of progress testing, we will discuss these issues and some attempts to solve them.
Arno Muijtjens was trained in Electrical Engineering (Measurement and Control) at the Eindhoven Technical University and has a PhD in Medical Informatics from Maastricht University. In 1999 he joined the Dept. of Educational Development and Research with a major interest in methodology and statistics in educational research, educational measurement, and assessment. More specific areas of interest are progress testing and computer based testing.