Correct answer: triangulation Using several data sources and checking whether findings converge is triangulation.

• randomization Randomization assigns participants to groups by chance.
• statistical control Statistical control adjusts for other variables in analysis.
• test equating Test equating makes scores from different test forms comparable.

Correct answer: quasi-experimental Intact classes, pre- and post-tests, and no random assignment describe a quasi-experimental design.

• true experimental A true experiment requires random assignment of learners.
• cross-sectional Cross-sectional studies collect data at one point in time.
• ethnographic Ethnographic research studies a culture through extended observation.

Correct answer: Simple random sampling Randomly selecting names from the complete list of the population is simple random sampling.

• Stratified random sampling Stratified sampling first divides the population into subgroups.
• One-stage cluster sampling Cluster sampling selects whole groups, not individuals.
• Systematic random sampling Systematic sampling selects at a fixed interval from the list.

Correct answer: The distractor is probably non-functioning. A distractor chosen by almost no examinees is not attracting anyone, so it is probably non-functioning.

• The keyed answer is probably incorrect. Rarely chosen distractors say nothing that suggests the keyed answer is wrong.
• The item therefore has strong discrimination. A non-functioning distractor does not by itself mean strong discrimination.
• The item therefore has strong construct validity. A non-functioning distractor does not by itself show construct validity.

Correct answer: About 84% of the reference group scored no higher. The 84th percentile means that about 84% of the reference group scored at or below the learner.

• The learner answered exactly 84% of the items correctly. Percentile rank is not the percentage of items answered correctly.
• The learner's raw score is 84 points above the mean. A percentile is not a number of points above the mean.
• Exactly 84% of all test items were easier for the learner. A percentile does not describe how many items were easier.

Correct answer: Cohen's kappa Cohen’s kappa measures agreement between two raters on categories after correcting for chance agreement.

• Pearson correlation Pearson correlation is for continuous variables and does not correct for chance agreement.
• Cronbach's alpha Cronbach’s alpha measures internal consistency of items.
• Spearman correlation Spearman correlation measures rank association, not categorical agreement.

Correct answer: Use a test blueprint with expert review. A test blueprint (table of specifications) with expert review shows that the test samples the syllabus and intended cognitive demands, supporting content validity.

• Compare scores with an external achievement criterion. Comparing scores with an external criterion is evidence of criterion validity.
• Estimate internal consistency across test items. Internal consistency is evidence of reliability, not of syllabus coverage.
• Convert raw scores into percentile ranks. Percentile ranks are a way of interpreting scores, not evidence of coverage.

Correct answer: 64.1-75.9 70 ± 1.96 × 3 = 70 ± 5.88, which gives about 64.1 to 75.9.

• 61.0-79.0 61.0–79.0 uses about ±3 SEM (±9).
• 67.0-73.0 67.0–73.0 uses ±1 SEM (±3).
• 68.5-71.5 68.5–71.5 uses ±0.5 SEM (±1.5).

Correct answer: -0.15 D = p_U − p_L = 0.70 − 0.85 = −0.15, a negative discrimination (more lower-group than upper-group learners got it right).

• 0.15 +0.15 has the wrong sign.
• 0.85 0.85 is the lower-group proportion, not the index.
• -0.85 −0.85 is not the difference between 0.70 and 0.85.

Correct answer: 0.75 The item difficulty index p is the proportion answering correctly: 75 ÷ 100 = 0.75.

• 0.25 0.25 is the proportion who answered incorrectly.
• 1.33 1.33 is 100 ÷ 75, the inverse of the proportion.
• 75 75 is the number (or percentage) correct, not the index.

Correct answer: Test-retest reliability Noticeably different scores for the same learners on a stable trait under comparable conditions a few days apart indicate weak test–retest reliability.

• Content validity Content validity concerns how well the items cover the content.
• Criterion validity Criterion validity concerns the relationship with an external criterion.
• Scoring objectivity Scoring objectivity concerns agreement among scorers, not changes in scores over days.

Correct answer: collect antecedent and consequence patterns before targeting intervention Collecting antecedent and consequence patterns (a functional approach) shows what is maintaining the behaviour before an intervention is chosen.

• assume that motivation is poor across all subjects Assuming poor motivation across all subjects is premature without evidence.
• remove preferred activities for the remainder of the week Removing preferred activities for a week is a heavy consequence without evidence of a pattern.
• raise task difficulty in every subject immediately Raising task difficulty in every subject could make matters worse.

Correct answer: Use wait time with structured response distribution. Wait time with structured response distribution gives more learners time and opportunity to answer while keeping questions challenging.

• Accept answers only from volunteers. Accepting only volunteers favours the confident learners who already dominate.
• Reduce questioning and rely mainly on explanation. Reducing questioning removes opportunities to participate.
• Award marks primarily for speed of response. Awarding marks for speed favours the fastest responders.

Correct answer: peer attention The peers’ laughter provides attention, and the behaviour drops when that reaction is prevented, which points to peer attention as the maintaining consequence.

• escape from academic demands Escape from academic demands would not depend on peer reactions.
• access to preferred materials Access to preferred materials would not depend on peer reactions.
• automatic sensory stimulation Automatic sensory stimulation would continue regardless of peer reactions.

Correct answer: Clustered tables with members facing one another Clustered tables with members facing one another support face-to-face discussion and shared materials.

• Straight rows facing the front Straight rows face the front and limit group talk.
• A horseshoe facing only the teacher A horseshoe facing only the teacher limits group discussion.
• Random seats changed during every activity Random seats changed every activity undermine group routines.

Correct answer: Establish accountability, repair harm and plan re-entry. A restorative approach establishes accountability, repairs harm, and plans re-entry.

• Keep the learners permanently separated without discussion. Permanent separation without discussion does not repair harm.
• Increase the penalty until neither learner wishes to speak. Increasing penalties is punitive rather than restorative.
• Ignore the incident once ordinary lessons resume. Ignoring the incident leaves the harm unaddressed.

Correct answer: You began promptly and checked each step carefully. “You began promptly and checked each step carefully” names exactly what the learner did, which is behaviour-specific praise.

• You are clearly the most talented learner in the group. Praising talent focuses on the person rather than a behaviour.
• This whole class has been generally good today. Generic praise for the whole class does not identify a behaviour.
• Well done, whatever it was that you were doing. “Whatever it was” names no behaviour at all.

Correct answer: Use proximity or a private cue before escalation. Proximity or a private cue is a minimal, non-disruptive first response to low-level off-task behaviour.

• Stop the class and issue a public warning. A public warning disrupts the whole class for a minor first event.
• Remove the learner immediately from the room. Removal from the room is excessive for a first, minor event.
• Assign a written punishment before speaking privately. Assigning a written punishment before speaking privately is disproportionate.

Correct answer: Teach, practise and cue a brief transition routine. Teaching, practising, and cueing a brief transition routine makes transitions quicker and more predictable.

• Increase homework until learners become more organised. More homework does not teach a transition routine.
• Change the transition procedure every week. Changing the routine weekly prevents it from becoming habitual.
• Wait for learners to discover a routine independently. Waiting for learners to discover a routine wastes lesson time.

Correct answer: refine conceptual reasoning through peer discussion Answering, discussing with peers, and answering again (peer instruction) is meant to refine conceptual reasoning.

• secure unrestricted consensus before marking The aim is not to force unrestricted consensus before marking.
• reward the fastest answer in each group The aim is not to reward the fastest answer.
• eliminate independent thinking before discussion Individual thinking comes first and is not removed.

Correct answer: differentiated instruction Keeping the same outcome while giving temporary supports such as sentence starters and organisers is differentiated instruction.

• ability tracking Ability tracking places learners in fixed groups or classes by ability.
• curriculum replacement Curriculum replacement changes the learning outcome itself.
• grade acceleration Grade acceleration moves a learner ahead to a higher grade.

Correct answer: predict-observe-explain Predicting, observing, and then explaining is the predict–observe–explain (POE) strategy.

• question-answer review A question-and-answer review checks recall after teaching and lacks the prediction step.
• read-copy-recall Read–copy–recall is a memorisation routine.
• listen-repeat-practise Listen–repeat–practise is a rehearsal routine.

Correct answer: Vary length, control other conditions and repeat trials. A fair test varies one factor (length), controls other conditions, and repeats trials to improve reliability.

• Change both length and mass during each trial. Changing both length and mass confounds the result.
• Conduct one trial for each selected length. One trial per length gives unreliable results.
• Change the release angle with every new length. Changing the release angle with each length also confounds the result.

Correct answer: reciprocal teaching Predicting, questioning, clarifying, and summarising while gradually taking over the roles is reciprocal teaching.

• concept mapping Concept mapping builds diagrams of relationships between ideas.
• programmed instruction Programmed instruction uses stepwise, self-paced materials.
• didactic teaching Didactic teaching is teacher-led transmission of content.

Correct answer: Model explicitly, guide practice, then supervise independence. For novices learning a safety-critical procedure, model explicitly, guide practice, then supervise independence (gradual release).

• Begin with independent performance before any demonstration. Independent performance before any demonstration is unsafe for novices.
• Use open exploration and discuss safety after errors occur. Discussing safety only after errors is unacceptable in a safety-critical task.
• Rely on peer discovery without modelling the procedure. Peer discovery without modelling risks errors and accidents.