Measurement, Assessment and Evaluation; Formative vs Summative Assessment

• Test – an instrument or set of questions used to obtain a sample of behaviour.
• Measurement – assigning numbers to a characteristic (e.g. 34/50). It is quantitative.
• Assessment – the broad, continuous process of collecting information about learning using many tools (tests, observation, projects, portfolios, questioning).
• Evaluation – making a value judgement about the quality or worth of learning or a programme, based on assessment information (e.g. "the student has met the standard", "the programme was effective").

Relationship: Test ⊂ Measurement ⊂ Assessment → used for Evaluation. Evaluation = quantitative description + qualitative description + value judgement.

• Placement – before instruction, to place students at the right level.
• Diagnostic – identifies specific learning difficulties and their causes (before or during instruction).
• Formative – during instruction, to monitor progress and improve teaching and learning. Low-stakes; feedback is the key. Examples: oral questioning, exit tickets, quizzes, thumbs up/down, observation, peer review of drafts.
• Summative – at the end of a unit/term, to judge achievement and give grades or certificates. High-stakes. Examples: final examinations, board exams, end-of-unit tests.

The terms formative and summative were introduced by Michael Scriven (1967) for programme evaluation and applied to student learning by Benjamin Bloom.

Assessment for learning (AfL) – formative; information is used to adjust teaching (Black & Wiliam, Inside the Black Box, 1998).
Assessment of learning – summative; certifies achievement.
Assessment as learning – students monitor their own learning through self-assessment and reflection (metacognition).

• Norm-referenced – a student's performance is compared with that of other students (percentile, rank, position in class). Useful for selection.
• Criterion-referenced – performance is compared with a fixed standard or learning objective (e.g. "can solve 8 out of 10 two-step word problems"). Useful for mastery and diagnosis. Term popularised by Robert Glaser (1963).

A teaching licence test with a fixed pass mark is closer to criterion-referenced interpretation.

Feedback is information about the gap between current performance and the goal, and how to close it. Effective feedback is:

• Specific – "Your topic sentence is clear; add two examples to support it" rather than "Good".
• Timely – given while the learning is still fresh.
• Focused on the task and process, not on the person ("You are lazy").
• Actionable – the student knows what to do next.
• Manageable – a few key points, not every error.
• Two-way – students respond to it (e.g. correction time).

Comment-only marking can be more effective for improvement than grades alone, because students attend to the comments.