Evaluation cycle

The evaluation cycle is the instructor's closed-loop procedure for knowing whether training is working and what to do when it is not. Evaluation is not a gate bolted onto the end of a course. It is continuous monitoring of objectives, trainee response, and the quality of the instruction itself. In airline training the correct philosophy is criterion-referenced testing (CRT): judge performance against job standards, not against the rest of the cohort.

Why evaluate

A trainer must always be able to answer:

  • is the course succeeding against its objectives?
  • if not, what corrective action is required?

Evaluation tells you how many trainees meet the objectives and whether the instruction or the test items need repair. A cohort that always fails the same item is as likely a design fault as a trainee fault.

Criterion-referenced versus normative

Approach Reference Question answered Fit for flight training
Normative referenced testing (NRT) Other students Who ranks higher? Poor: rank does not equal safe
Criterion-referenced testing (CRT) Job / syllabus standard Is this person good enough? Required: same stance as EBT grading

Profile reporting keeps CRT information usable: strengths and weaknesses against standards, not a single collapsed mark. Competency × grade word pictures are a profile system, not a league table. EBT competence decisions reference the approved programme standards only.

Three functions of evaluation

An assessment must declare which function it serves; validity does not transfer across functions.

Function Question Timing / action
Diagnostic What gaps exist so we can remediate? During training; must be followed by remedial teaching
Achievement Has the phase standard been met? End of phase; pass / remediate and retest
Predictive How will this person perform on the job later? Statistical design against job performance; not a relabelled achievement test

A diagnostic exam only at course end is a contradiction: diagnosis has training value only when there is still time to act.

Five-step cycle

  1. Define objectives. Syllabus contract. Fixes function (diagnostic / achievement / predictive) and makes validity checkable.
  2. Select the criteria. Test specification: coverage, conditions, item format, number of items vs time, weighting and marking. Written before administration, it becomes the audit trail.
  3. Choose the best method. Obey the fidelity principle: the method should best approximate the behaviour the objective names. Multiple-choice of a rejected-take-off decision is not the same as flying it in the FSTD.
  4. Administer the evaluation. Conditions match the plan; bias controlled as far as practical.
  5. Draw conclusions. Feed results to trainee remediation, instructor self-correction, and objective or item redesign. Close the loop.

Item quality bar

Every item should be valid, reliable, objective, differentiating, and comprehensive. Fail any one and the score measures the flaw more than the trainee.

How this sits beside EBT grading

Observe, record, classify, evaluate (ORCE) and the evidence-based training competency grading method (VENN) are the shop-floor realisation of continuous criterion-referenced assessment inside a module. Chapter 11 supplies the design philosophy. Competency frameworks supply the standards. Inter-rater reliability keeps grades usable as data. Do not invent a parallel NRT "who is best in this course" culture alongside EBT profiles.

Instructor use

  1. State the objective and function before writing any item or scenario grade plan.
  2. Prefer CRT profiles over single ranks when reporting to the next instructor or line training.
  3. Match method fidelity to the behaviour: knowledge → written/oral; handling → FSTD/aircraft; behaviour → observed scenario performance.
  4. After diagnostic evidence, teach the gap the same day if possible.
  5. When a whole group fails one item, inspect the item and the teaching before blaming the group.
  6. Align oral and written questions with Bloom level of the objective.

Connections

Sources