Fishbone Analysis
A collaborative method for organizing possible causes, identifying evidence needs, and moving from symptoms toward testable explanations.
Also called a cause-and-effect or Ishikawa diagram, a fishbone diagram helps a team identify and organize many possible causes of a defined outcome. It supports root-cause investigation, but the diagram alone does not prove that any proposed cause is true.
How a fishbone diagram works
The outcome or problem statement forms the “head” of the fish. A horizontal line forms the spine. Major cause categories branch from the spine, and increasingly specific possible causes branch from those categories as smaller bones.
Knowledge, roles, communication, staffing, participation
Instruction, workflows, routines, sequencing, implementation
Curriculum, technology, data, assessments, resources
Time, schedules, climate, facilities, external conditions
Rules, priorities, incentives, accountability, decision rights
Definitions, data quality, timing, comparability, interpretation
Defined effect or outcome
Specific, observable, measurable, and neutral
Seven steps for conducting the analysis
Define the effect
Write a focused, evidence-based statement describing what is happening, where, for whom, and over what period. Avoid embedding an assumed cause or preferred solution.
Assemble relevant perspectives
Include people who understand the work and those most affected by it. Establish norms that make it safe to examine systems without blaming individuals or groups.
Choose useful categories
Select five to seven category prompts suited to the issue. People, methods, tools, materials, environment, policy, and measurement are common starting points.
Generate possible causes
Brainstorm broadly before evaluating ideas. Place each possible cause under the most useful category; duplicate an idea when it genuinely connects to more than one area.
Develop subcauses
Ask “Why might this occur?” until the team reaches a level specific enough to investigate or influence. Distinguish observable conditions from judgments about people.
Validate with evidence
Identify what data, observation, document review, interview, or process study could confirm or challenge each high-priority hypothesis. Look for disconfirming evidence.
Prioritize and test action
Select causes supported by evidence and within the team’s influence. Design a small, measurable change, identify ownership, monitor intended and unintended effects, and revise the diagram as the team learns.
School-district example: mathematics achievement
Example effect statement
During the past two assessment cycles, the percentage of students meeting the district’s grade-level mathematics benchmark declined in grades six through eight, with different patterns across schools and student groups.
The following are hypotheses to investigate, not conclusions:
Instruction and learning
- Uneven implementation of adopted instructional materials
- Limited opportunities for mathematical discourse and problem solving
- Scaffolds that do not consistently preserve grade-level reasoning
- Insufficient response to formative evidence during instruction
People and professional learning
- Variation in content-specific preparation or coaching access
- Limited collaborative planning and student-work analysis
- Unclear roles across teachers, interventionists, and specialists
- Student or family perspectives not represented in improvement planning
Curriculum, tools, and resources
- Supplemental resources are not aligned with adopted standards or materials
- Students have inconsistent access to manipulatives or accessible formats
- Technology is used for task completion without strengthening reasoning
- Intervention materials do not connect coherently to core instruction
Time, environment, and opportunity
- Interrupted instructional time or uneven course placement
- Attendance patterns reduce access to essential learning sequences
- Schedules limit intervention, acceleration, or teacher collaboration
- Classroom participation structures do not engage every learner
Policy and implementation
- Pacing expectations conflict with time needed for student understanding
- School improvement priorities or resource allocations are inconsistent
- Course-placement or intervention criteria are unclear
- Implementation expectations are not paired with usable support
Measurement and interpretation
- Benchmark versions, administration conditions, or participation changed
- Aggregate results conceal differences among standards or student groups
- Measures emphasize procedural fluency more than conceptual understanding
- Data arrive too late or lack the detail needed for instructional response
Facilitation guardrails
- Describe systems and conditions—not character
Avoid labels such as “unmotivated students,” “resistant teachers,” or “uninvolved families.” Translate judgments into observable conditions, experiences, processes, and evidence that can be investigated.
- Do not confuse brainstorming with proof
Items on the diagram are hypotheses. The number of sticky notes, strength of opinion, or position of the speaker does not establish causation.
- Include people closest to the experience
Students, families, classroom educators, support staff, and operational teams may see conditions that are invisible in aggregate data or leadership discussions.
- Look for interactions among causes
Complex outcomes rarely have a single cause. Use additional tools—such as process maps, Pareto charts, interviews, or relations diagrams—when the connections among factors matter.
Templates and additional guidance
SDLA fishbone protocol
Download the printable protocol and diagram template for a facilitated team session.
Download the templateASQ fishbone guidance
Review the American Society for Quality’s overview, procedure, examples, and related quality-improvement tools.
Review ASQ guidanceIHI cause-and-effect tool
Explore an additional overview, instructions, and template from the Institute for Healthcare Improvement.
Open the IHI resource