Study guide

18.4 Standard Setting and the Angoff Calculation

BCSP CIT Certified Instructional Trainer Study Guide 2026

Track your progress as you read

Finished sections get a check mark.

1 Credential and Training Standards

1.1 BCSP CIT Credential Scope, Exam Logistics & Recertification 1.2 ANSI/ASSP Z490 Criteria for Accepted EHS Training Practices

2 Adult Learning and Learning Access

2.1 Foundations of Andragogy: Knowles' Adult Learning Assumptions in Safety 2.2 Learner Literacy, ESL Accommodations & Plain Language Compliance 2.3 Opening a Course: Administrative Items and Learning Expectations

3 Communication and Facilitation

3.1 Verbal & Nonverbal Facilitation Techniques in Safety Training 3.2 Active Listening, Socratic Questioning & Teach-Back Verification 3.3 Feedback Frameworks, Corrective Coaching & Psychological Safety

4 Conflict and Instructional Technology

4.1 Managing Challenging Classroom Dynamics, Resistance & Conflict 4.2 Instructional Media, Presentation Tools & Technology Contingencies

5 Performance Needs and Existing Training

5.1 Performance Gaps and Training Necessity 5.2 Evaluating Existing Training and Interventions

6 Organizational, Task and Hazard Analysis

6.1 Organizational and Job Task Analysis 6.2 Hazard Analysis, KSAs and Task Evidence

7 Audience, Baseline and Prerequisites

7.1 Defining the Audience and Its Learning Needs 7.2 Baseline Knowledge and Prior-Learning Credit 7.3 Verifying Prerequisites and Readiness for Training

8 Applicable Training Requirements

8.1 OSHA Training Frameworks, General Duty Clause & Consensus Standards 8.2 Mandated Safety Programs: Lockout/Tagout, Confined Space & Fall Protection 8.3 Chemical & Biological Hazards: HazCom, Bloodborne & Respiratory Protection

9 Systematic and Iterative Design

9.1 ADDIE and the Systematic Approach to Training 9.2 SAM, Agile Development and Prototypes

10 Objectives and Learning Outcomes

10.1 Writing Observable ABCD Objectives 10.2 SMART, Objective Hierarchies and Stakeholder Consensus

11 Learning Domains and Practice

11.1 Cognitive Learning and Revised Bloom Taxonomy 11.2 Psychomotor, Affective and Constructive Alignment 11.3 Experiential Learning Cycles (Kolb), Adult Discovery & Instructional Scaffolding

12 Delivery Choices and Program Economics

12.1 Modality Matrix: Classroom ILT, E-Learning, Microlearning, Hands-On & OJT 12.2 Cost Classification and Worked Budgets 12.3 Resource Feasibility and Cost-Benefit Decisions

13 Developing and Validating Course Components

13.1 Lesson Plan Architecture: Components of Comprehensive Trainer & Participant Guides 13.2 Job Aids and Performance Support 13.3 SME Selection, Credible Sources and Validation

14 Development Planning and External Providers

14.1 Implementing the Instructional Development Plan 14.2 Deciding When and How to Outsource Development

15 Pilots and Recoverable Records

15.1 Representative Pilots and Course Revision 15.2 Record Retention, Privacy and Recovery

16 Implementation Environments and Logistics

16.1 Physical Classroom Design, Seating Arrangements, Ergonomics & Environmental Controls 16.2 Field, Shop & Hands-On Training: Safety Briefings, PPE Verification & Live Demonstration Safety 16.3 Virtual Instructor-Led Training (VILT), Technical Contingency Planning & Remote Engagement 16.4 Scheduling, Logistics and Course Time Management

17 Assessment and Remediation

17.1 Formative and Summative Assessment Tools 17.2 Assessment Criteria, Alignment and Remediation

18 Items, Performance and Standards

18.1 Writing Clear Items and Plausible Distractors 18.2 Item Analysis and Assessment Data 18.3 Performance Checklists and Evaluator Calibration 18.4 Standard Setting and the Angoff Calculation

19 Evaluation Models, Surveys and Learning Data

19.1 Evaluation Models and Feedback Surveys 19.2 Learning Outcomes and Evaluation Data

20 Transfer and Behavior Evidence

20.1 Transfer of Learning and Work Conditions 20.2 Behavior Observations and Change Reporting

21 Results, ROI and Improvement Reports

21.1 Organizational Results and Causal Limits 21.2 ROI, Course Improvement and Training Reports

18.4 Standard Setting and the Angoff Calculation

Key Takeaways Explain This Topic AI

Angoff panel ratings estimate a minimally competent candidate's item success probability.

Thresholds need a documented candidate definition, evidence and decision rule.

The five-item example estimates 3.96 correct; an at-least rule requires four whole correct items.

Last updated: October 2026

Standard-Setting Methodologies & The Modified Angoff Walkthrough

A passing threshold should be supported by the purpose and performance requirements, rather than selected only because a familiar percentage is convenient. Employment-testing requirements depend on the use and context; they do not mandate the Angoff method for every workplace course.

The Modified Angoff method is one criterion-referenced standard-setting approach. A panel estimates the probability that a minimally competent candidate would answer each knowledge item correctly. Panel judgments require a clear competency definition, evidence and review; the method does not guarantee legal defensibility.

The Minimally Competent Candidate (MCC) Concept

The Angoff process hinges upon the concept of the Minimally Competent Candidate (MCC) (also known as the "borderline candidate"). The MCC represents a hypothetical trainee who possesses just enough knowledge, judgment, and skill to perform the job safely and effectively without endangering themselves or others. They are neither a master technician nor a reckless novice; they represent the threshold of acceptable professional competence.

Standard-setting workflow

Define the minimally competent candidate, recruit an appropriate panel, train members on the task, obtain independent ratings, discuss discrepancies and review evidence. Aggregate ratings according to the chosen method and document the final decision. Panel size and rounds depend on the assessment; a fixed six-to-twelve-person rule is not universal.

Angoff Standard-Setting Calculation Walkthrough

  • Knowledge item SME 1 SME 2 SME 3 SME 4 SME 5 Mean
  • 1 0.85 0.90 0.80 0.85 0.90 0.86
  • 2 0.95 0.95 0.90 0.90 0.95 0.93
  • 3 0.70 0.65 0.75 0.70 0.70 0.70
  • 4 0.65 0.70 0.60 0.65 0.65 0.65
  • 5 0.80 0.85 0.80 0.80 0.85 0.82

The five item means sum to 3.96, or 79.2% of five items. If a policy requires at least this many correct whole items, the threshold is four correct answers out of five (80%), because 3.96 items cannot be earned on a dichotomously scored test. This is a knowledge-test illustration, not a method for averaging away critical performance failures.

Retesting Protocols and Record Integrity

Provide targeted remediation and an appropriate reassessment under the applicable program. Define retest conditions, evaluator responsibilities and restrictions before administration. Document the evidence and decision. Do not invent a universal prohibition on same-day retesting or treat a written retest as proof of a missing physical skill.

Reviewing a standard-setting decision

Panel ratings are judgments about a defined minimally competent candidate, not the observed percentage correct from one class. Define that candidate's expected training, duties and performance carefully. Panelists who imagine very different audiences can produce disagreement that averaging will conceal. Discussion should identify those different assumptions before the final decision.

Consider what the threshold permits. A five-item knowledge test offers only six possible raw scores: zero through five. Converting an estimate of 3.96 into a whole-item rule requires an explicit policy; four correct is the first whole score meeting at least 3.96. A percentage displayed with decimal precision does not create fractional credit where every item is dichotomous.

  • Standard-setting element Documentation
  • Candidate definition Expected minimum competence and conditions
  • Item judgments Independent estimates and review process
  • Evidence Content review and relevant performance data
  • Decision Final threshold, rounding rule and rationale

Standard setting cannot rescue an invalid assessment. If the items do not represent the objectives, a carefully calculated threshold still measures the wrong content. If the key is incorrect, the score is distorted. If a physical skill is assessed only by written recall, the instrument is insufficient for that outcome. Validate content and administration before interpreting the cut score.

Different purposes may call for different standard-setting approaches. A required procedural criterion may be directly specified by a controlling rule or approved task standard. A panel-estimation method is a useful concept for knowledge testing, but must not average away a safety-critical action. Keep these decision rules distinct.

After use, review the assessment's evidence and decisions. An unexpected failure pattern can reflect a true learning gap, a poor criterion, an item defect or a new audience. Investigate and document changes. Do not lower a threshold automatically because a class dislikes its results, or raise it to create a desired failure quota. The threshold should remain connected to the required competence.

Avoiding false precision

A panel mean such as 79.2% can look more precise than the judgment process supports. Record assumptions, disagreement and the final decision rule. Do not imply that the decimal alone proves a scientifically exact competence boundary.

Standard-setting documentation helps a later reviewer understand how the threshold was chosen and whether changed objectives require reconsideration. It does not exempt the team from validating the items, administration and scoring. The entire evidence chain matters for a sound decision.

Separating threshold review from learner remediation

If many learners narrowly miss a threshold, first examine the evidence. Was the task taught and practiced appropriately? Were the items accurate and representative? Were conditions comparable and accessible? Was scoring correct? An unexpected result warrants investigation, not an automatic threshold change. A confirmed instrument defect needs fair correction; a genuine learning gap needs remediation.

Keep the decision rationale distinct from the remediation record. The standard-setting file explains what level of evidence is required. The learner record explains which criteria were met, which were not and what happened afterward. Mixing these can make it appear that a threshold was changed to pass a particular person.

For performance assessment, the approved critical-step rule remains separate from the knowledge-test calculation. A person scoring four of five on the illustrative written test may meet its raw-score threshold while still lacking a required physical check. Conversely, a successful observed task does not necessarily answer every knowledge requirement. Use the evidence required by the program rather than substituting one favorable measure for another.

Review the standard when the role, objective or criterion changes. Preserve earlier versions and the reason for revision under the record policy. This lets future reviewers understand which standard applied to a particular decision and avoids applying a new threshold retroactively without an appropriate process.

Key takeaways

Angoff panel ratings estimate a minimally competent candidate's item success probability.

Thresholds need a documented candidate definition, evidence and decision rule.

The five-item example estimates 3.96 correct; an at-least rule requires four whole correct items.