QLS cheat sheet

Behaviorism: learning = observable behavior change produced by stimulus-response-reinforcement contingencies; measurable outcomes drive design.

Cognitivism: learning = organized mental schemas; instruction should manage cognitive load and support encoding, storage, and retrieval.

Constructivism: learners build knowledge from prior experience; design favors authentic tasks, scaffolding, and social negotiation over transmission.

Gagne's nine events: gain attention, state objectives, recall prior knowledge, present content, guide learning, elicit performance, give feedback, assess, enhance retention/transfer.

Bloom's revised taxonomy (ascending): remember, understand, apply, analyze, evaluate, create; verbs map to assessable cognitive demand.

Merrill's First Principles: activation, demonstration, application, integration - and task-centered sequencing of instruction.

Cognitive Load Theory: working memory is limited; split-attention and redundancy effects raise extraneous load and depress learning.

ADDIE phases run analysis, design, development, implementation, evaluation; evaluation feeds back into analysis in iterative cycles.

Backward design order: identify desired results, determine acceptable evidence, then plan learning experiences and instruction.

Constructive alignment: learning outcomes, teaching/learning activities, and assessment tasks must target the same verbs and content.

Learning objective formula: performance (observable verb) + condition (given what) + criterion (standard of acceptable performance).

Task analysis decomposes a terminal objective into prerequisite sub-skills; prerequisites are sequenced before dependent skills.

Scope = breadth and depth of content; sequence = order of delivery; both derive from learner entry level and task dependencies.

Chunking rule: segment content into modules sized to working-memory limits, with practice between chunks rather than massed at the end.

Content authoring standards: single-source reusable objects tagged with metadata support reuse across courses and platforms.

Questioning wait time: 3-5 seconds after a prompt increases response length and participation breadth versus calling immediately.

Think-pair-share sequence: individual reflection, paired discussion, whole-group synthesis - raises participation over cold calling.

Feedback timing: immediate feedback aids skill acquisition; delayed feedback can aid retention and transfer for complex tasks.

Group size trade-off: smaller groups raise individual airtime; larger groups raise idea diversity but reduce per-learner speaking time.

Psychological safety (Edmondson): learners speak up when interpersonal risk is low; leader inclusiveness and framing of errors as data raise it.

Facilitator role shifts along a continuum from direct instruction to coaching to observation as learner expertise increases.

Nominal group technique: silent idea generation, round-robin recording, clarification, then ranked voting to reduce dominance effects.

Validity = evidence that scores support the intended interpretation; reliability = consistency of scores; reliability is necessary but not sufficient for validity.

Criterion-referenced assessment compares performance to a standard; norm-referenced compares learners to each other.

Formative assessment informs learning during instruction; summative assessment certifies achievement after instruction.

Kirkpatrick levels: reaction, learning, behavior, results - each level measures a different outcome and does not guarantee the next.

Item difficulty index p = proportion answering correctly; values near 0 or 1 provide little discrimination.

Item discrimination index D = p(upper group) − p(lower group); values below ~0.20 flag weak items for review.

Criterion-referenced cut score should be set by a documented method (e.g., modified Angoff) with subject-matter expert judgment.

Inter-rater reliability is estimated by percent agreement or a kappa/ICC statistic; low agreement undermines score validity.

SCORM packages communicate runtime data to an LMS; xAPI (Tin Can) records statements of actor-verb-object outside the LMS.

LTI standardizes tool launch and grade passback between an LMS and external tools without custom integrations.

WCAG conformance levels ascend A, AA, AAA; most procurement and legal baselines target AA.

Asynchronous tools decouple time and place; synchronous tools require co-presence but enable immediate feedback.

Blended learning combines face-to-face and online delivery; the rotation model cycles learners through stations including online learning.

Learning analytics uses logged interaction data to predict and intervene; privacy and consent govern what may be collected.

Data interoperability: CSV imports, APIs, and standards-based exports determine whether records survive a platform change.

Perceivable, Operable, Understandable, Robust (POUR) are the four WCAG principles; each success criterion maps to one.

Alternatives rule: non-text content needs text alternatives; prerecorded video needs captions; audio needs transcripts.

Contrast minimum: normal text requires 4.5:1 against background; large text (18pt or 14pt bold) requires 3:1.

Never convey meaning by color alone; pair color with text, pattern, or icon to satisfy use-of-color criteria.

Universal Design for Learning: multiple means of engagement, representation, and action/expression.

Accommodations change how a learner demonstrates knowledge (e.g., extended time) without altering the construct being measured.

Assistive technology stack: screen readers rely on semantic structure and headings; keyboard-only access requires visible focus and logical tab order.

Informed consent requires disclosure of purpose, risks, benefits, voluntariness, and the right to withdraw without penalty.

Confidentiality rule: collect the minimum learner data needed; share only with those who have a legitimate need to know.

Copyright: fair-use analysis weighs purpose, nature, amount, and market effect; open licenses (e.g., CC BY) pre-authorize reuse under stated terms.

Attribution: CC BY requires credit, license link, and indication of changes; CC BY-SA adds share-alike obligations.

Assessment security: item exposure, answer sharing, and proctoring practices must be disclosed to learners in advance.

Conflict of interest: disclose any financial or personal interest that could bias design, delivery, or evaluation decisions.

Records retention and privacy obligations follow applicable law and institutional policy; destroy data when the stated purpose ends.

Triple constraint: scope, schedule, and cost are interdependent; a change to one requires renegotiating the others.

Critical path = longest dependent sequence of tasks; it sets the minimum project duration and has zero float.

RACI matrix assigns each deliverable a Responsible, Accountable, Consulted, and Informed party; exactly one Accountable per item.

Stakeholder power/interest grid: high power-high interest parties get manage closely; low power-low interest get monitor.

Status reporting rule: report against a baseline and state variance in scope, schedule, and budget with corrective actions.

Risk register fields: description, likelihood, impact, score, owner, response (avoid, mitigate, transfer, accept).

Change control: documented request, impact assessment, approval authority, then baseline update - undocumented changes erode control.

Behaviorism: learning = observable behavior change from stimulus-response-reinforcement contingencies; design around measurable outcomes.

Cognitivism: learning = organized mental schemas; instruction manages cognitive load and supports encoding, storage, retrieval.

Constructivism: learners build knowledge from prior experience; favor authentic tasks, scaffolding, and social negotiation.

Gagne's nine events: gain attention, state objectives, recall prior knowledge, present content, guide learning, elicit performance, give feedback, assess, enhance transfer.

Bloom's revised taxonomy ascending: remember, understand, apply, analyze, evaluate, create; verbs signal assessable cognitive demand.

Merrill's First Principles: activation, demonstration, application, integration, sequenced around a whole task.

Cognitive Load Theory: working memory is limited; split-attention and redundancy effects raise extraneous load and depress learning.

ADDIE runs analysis, design, development, implementation, evaluation; evaluation feeds back into analysis on each cycle.

Backward design order: identify desired results, determine acceptable evidence, then plan learning experiences and instruction.

Constructive alignment: outcomes, activities, and assessments must target the same verbs and content.

Objective formula: performance (observable verb) + condition (given what) + criterion (standard of acceptable performance).

Task analysis decomposes a terminal objective into prerequisite sub-skills; prerequisites are sequenced before dependent skills.

Scope = breadth and depth of content; sequence = delivery order; both derive from entry level and task dependencies.

Chunking rule: segment content to working-memory limits and place practice between chunks rather than massed at the end.

Reusable learning objects carry metadata tags so content can be single-sourced and reused across courses and platforms.

Wait time of 3-5 seconds after a prompt increases response length and participation breadth versus calling immediately.

Think-pair-share: individual reflection, paired discussion, whole-group synthesis - raises participation over cold calling.

Immediate feedback aids skill acquisition; delayed feedback can aid retention and transfer on complex tasks.

Smaller groups raise individual airtime; larger groups raise idea diversity but cut per-learner speaking time.

Psychological safety (Edmondson): learners speak up when interpersonal risk is low; framing errors as data raises it.

Facilitator role shifts from direct instruction to coaching to observation as learner expertise increases.

Nominal group technique: silent idea generation, round-robin recording, clarification, then ranked voting to reduce dominance.

Validity = evidence scores support the intended interpretation; reliability = score consistency; reliability is necessary but not sufficient for validity.

Criterion-referenced assessment compares performance to a standard; norm-referenced compares learners to each other.

Formative assessment informs learning during instruction; summative assessment certifies achievement after instruction.

Kirkpatrick levels: reaction, learning, behavior, results - each measures a different outcome and does not guarantee the next.

Item difficulty p = proportion answering correctly; values near 0 or 1 discriminate poorly.

Item discrimination D = p(upper) − p(lower); values below about 0.20 flag weak items for review.

Cut scores should be set by a documented method (e.g., modified Angoff) using subject-matter expert judgment.

Inter-rater reliability is estimated by percent agreement or kappa/ICC; low agreement undermines score validity.

SCORM packages exchange runtime data with an LMS; xAPI records actor-verb-object statements outside the LMS.

LTI standardizes tool launch and grade passback between an LMS and external tools without custom integrations.

WCAG conformance levels ascend A, AA, AAA; most procurement and legal baselines target AA.

Asynchronous tools decouple time and place; synchronous tools require co-presence but enable immediate feedback.

Blended learning combines face-to-face and online delivery; the rotation model cycles learners through stations including online learning.

Learning analytics uses logged interaction data to predict and intervene; privacy and consent govern what may be collected.

Data interoperability - CSV, APIs, standards-based exports - determines whether records survive a platform change.

POUR - Perceivable, Operable, Understandable, Robust - are the four WCAG principles; each success criterion maps to one.

Alternatives rule: non-text content needs text alternatives; prerecorded video needs captions; audio needs transcripts.

Contrast minimum: normal text 4.5:1 against background; large text (18pt or 14pt bold) 3:1.

Never convey meaning by color alone; pair color with text, pattern, or icon to satisfy use-of-color criteria.

Universal Design for Learning: multiple means of engagement, representation, and action/expression.

Accommodations change how a learner demonstrates knowledge (e.g., extended time) without altering the construct measured.

Screen readers depend on semantic structure and headings; keyboard-only access requires visible focus and logical tab order.

Informed consent requires disclosure of purpose, risks, benefits, voluntariness, and the right to withdraw without penalty.

Confidentiality rule: collect the minimum learner data needed and share only with those having a legitimate need to know.

Fair-use analysis weighs purpose, nature, amount, and market effect; open licenses such as CC BY pre-authorize reuse on stated terms.

CC BY requires credit, license link, and indication of changes; CC BY-SA adds share-alike obligations.

Assessment security: item exposure, answer sharing, and proctoring practices must be disclosed to learners in advance.

Conflict of interest: disclose any financial or personal interest that could bias design, delivery, or evaluation decisions.

Records retention and privacy follow applicable law and institutional policy; destroy data when the stated purpose ends.

Triple constraint: scope, schedule, and cost are interdependent; changing one requires renegotiating the others.

Critical path = longest dependent task sequence; it sets minimum project duration and has zero float.

RACI assigns each deliverable a Responsible, Accountable, Consulted, and Informed party; exactly one Accountable per item.

Power/interest grid: high power-high interest parties get manage closely; low power-low interest get monitor.

Status reporting rule: report against a baseline, state variance in scope, schedule, and budget, and name corrective actions.

Risk register fields: description, likelihood, impact, score, owner, response (avoid, mitigate, transfer, accept).

Change control: documented request, impact assessment, approval authority, then baseline update; undocumented changes erode control.