Criterion-referenced (CRA) and norm-referenced (NRA) assessment

Criterion-referenced assessment

In criterion-referenced assessment, students' performances are judged against pre-set criteria as specified in the intended learning outcomes. It informs teachers how well the intended learning outcomes have been achieved by students. For CRA to be beneficial to learning, explicit criteria are essential which provide clear learning goals to students.

Norm-referenced assessment

Norm-referenced assessment compares students with others. Students are assigned grades according to their standing relative to other students. Grades are commonly allocated in quotas that follow the normal curve (grading on the curve). NRA does not however say anything about the standard of students' performances, only about which students are better than others. For this reason, there are strong moves internationally to move towards CRA.

Qualitative and quantitative assessment

Qualitative assessment

Qualitative assessment is based on the standards model. It defines forms of knowledge to be reached at the end of teaching, expressed as various levels of acceptability in the objectives and grading system. Teachers have to judge what grading criterion best suits a given student's performance.

In a qualitative framework, learning is evaluated according to the qualities displayed in a student's performance. Learning builds upon previous knowledge and its structure becomes more complex. Assessment should therefore inform the present state of complexity, and how that matches the intended learning outcomes. The point of qualitative assessment is not how much the final grade is, but whether it tells us how well the performance matches the objectives. Learning outcomes should therefore be assessed holistically.

Quantitative assessment

Quantitative assessment is based on the measurement model, borrowed from individual differences psychology. NRA requires a quantitative framework for assessment so that students may be compared. It is therefore necessary to reduce students' performances to a unidimensional scale, most frequently percentages, so that comparisons between students are possible. Learning is evaluated according to how much correct material has been learned and can be displayed, or is rated on arbitrary scales.

It is commonly assumed that percentages are a universal currency, equivalent across subject areas, and across students, so that different students' performances in different subjects can be directly compared. This is completely unsustainable - there is no way one can say for example that 58% in Physics 101 is equivalent to 58% on Textile Design 103. Yet such assumptions are made every day. Numbers provide a misleading 'objectivity' that does not in fact exist.

Declarative knowledge, functioning knowledge and performance assessment

Declarative knowledge

Declarative Knowledge refers to knowing about things. It is public knowledge, subject to rules of evidence that make it verifiable, replicable and logically consistent. It is what is in libraries, textbooks and research papers, and is what teachers "declare" in lectures. Most common assessment techniques used in universities are aiming at assessing declarative knowledge.

Functioning knowledge

Functioning knowledge is based on the idea that true understanding changes the way that students actually function in their discipline areas. The knowledge allows them to integrate academic knowledge with the skills required for that profession and the context for using them to solve problems. This knowledge is within the experience of the learners, who can now put declarative knowledge to work by solving real life problems in the professions for which they are being prepared.

Performance assessment

In professional education, while declarative knowledge provides the conceptual foundation, students should be using and applying that knowledge not just talking or writing about it. It follows that assessment should therefore be directed at assessing functioning knowledge as close to its real-life context as possible. This is referred to as 'authentic assessment' or 'performance assessment'.

Formative and summative assessment

Formative assessment

Formative assessment provides informative feedback about how learning is proceeding, providing teachers and students with diagnostic information and guidance for improvement. Feedback in the form of comments given with summative assessment is too late as the learning in the unit in question is over. For suggestions on providing effective feedback, see feedback.


Summative assessment (grading)

Summative assessment is used to grade students at the end of a unit, or to accredit at the end of the programme. Its purpose is to see how well students have learned what they were supposed to have learned. See deriving grades.

Analytic and holistic assessment

Analytic assessment

Analytic assessment rates student performance on independent aspects of the task, not on the task as a whole. Usually, these aspects are weighted differentially to derive the final grade. The major disadvantages of analytic assessment are that it demands a quantitative method of assessment and that when the task is more than the sum of its parts, the assessment is deficient.

An analytic assessment could easily allow a student to pass by averaging performances on particular aspects of a complex task even if they failed one crucial aspect. However, who would like to be operated upon by a surgeon who had passed every aspect of performing an operation very well, but who failed on safety procedures? On a strict application of the analytical model, that student would have to pass! Analytic assessment is however essential when providing formative feedback. (See analytic rubrics).

Holistic assessment

Holistic assessment addresses the whole performance, not just aspects of it. Holistic assessment recognises the intrinsic meaning of the target performances and is therefore a truer method of assessing such performances. Most professional acts - problem solving, decision-making, assessing case studies - are holistic by nature. It makes little sense to assess students only on how well they carry out parts of the act if they are not assessed on how well they perform the act itself. (See holistic rubrics)

Assessing learning outcomes holistically needs a conceptual framework that enables one to judge the relationship between the parts and the whole. Such a framework is unique to each discipline, but a generic framework like the SOLO taxonomy can be useful in deriving a subject-specific framework.

Continuous/progressive and final/terminal assessment

Continuous / progressive assessment

Progressive or continuous assessment uses results taken during the subject while learning is proceeding, for grading purposes. Continuous assessment should only be used when assessing partial outcomes that need to be independently mastered early in the semester, or when assessing basic declarative knowledge. Continuous assessment also allows formative monitoring of students' progress by revealing errors so that it may be corrected.

Final / terminal assessment

Final or terminal assessment is carried out at the end of a subject or an educational programme. The student is required to sit an examination or submit a major item of work assignment to obtain a final grade. Traditionally, final assessment provides a summative evaluation on student performance and determines whether the student can proceed to the next level of study or be accredited. Final assessment takes into account all the content to be learned, but if all assessments are at the end of a semester there is a danger of stress and overload - for teachers as well!