When evaluating student progress through ICT-based activities, which method…

Concept: Reliability in assessment is the property of a measurement that yields the same picture of a learner's attainment each time it is repeated. Three conditions produce it:

• the same task or item set for every learner, so the performances being compared are comparable;

• a scoring rule fixed in advance and applied identically to every response, so the score does not move with the marker or with the moment of marking;

• a stored response record, so any score can be re-read and re-scored later.

Reliability drops whenever one of those conditions is carried by a human judgement made in real time. A rating formed while a child is speaking cannot be re-scored at all, and marking done by hand against a key still admits marker error and drift across a set of scripts. A system that applies a stored key to a stored response removes that source of variation, because the same response always earns the same score.

One further condition applies when the learning is carried by an ICT activity: the progress to be judged has been generated inside that activity, so an assessment running within it measures the very performance in question and captures the activity's own response data, while any assessment conducted outside it measures a separate performance and can only stand as a proxy.

Application: Here progress is generated through ICT-based activities in a primary English classroom, and the technique wanted is the one whose reading of that progress stays most consistent. A digital interactive quiz delivers the same item set to every child, scores each response against a stored key at the moment of submission, and keeps an item-level log per child that the teacher can reopen later. All three reliability conditions are met, the scoring is applied by the system rather than by a marker, and the measurement happens inside the ICT activity whose progress is being judged, so repeating it returns the same score for the same performance.

Contrast with the other techniques:

• Printed worksheets: the strongest near-miss here, because a worksheet also gives every child the same items and is marked against a key. Two things separate it on reliability: the marking is done by the teacher by hand, so marker error and drift across scripts enter the score; and what is recorded is a separate paper performance rather than the work done inside the ICT activity, so the activity's own item-level response data never reaches the assessment.

• Peer discussion: the prompts a child meets depend on the group, the talking time and the partners of the day, and the rating is formed from a live impression that leaves no re-scorable record, so a second administration seldom places the child at the same level.

• Oral questioning: the questions put to individual children usually differ and the reply is rated as it is spoken, so the score carries the judgement of that moment and no stored response remains to re-check.

Cross-check: Run the definition backwards as a check: would two separate administrations place the same child at the same level? Talk-based and on-the-spot techniques leave nothing re-scorable and shift with the group, the question and the moment; a hand-marked paper set keeps the items fixed but hands the scoring to a human marker and reads a performance produced outside the activity being assessed. The technique that keeps a fixed item set, a stored key applied automatically to a stored response, and does so inside the ICT activity itself, is the interactive quiz.

Result: The most reliable assessment of progress through ICT-based activities in a primary English classroom is interactive quizzes.