How to Calculate Adverse Impact
Calculate each group’s selection rate, divide each lower rate by the highest rate, and flag ratios below 0.80 for further review while considering sample size and statistical significance.
This guide turns the topic into a practical process rather than a list of generic tips. It explains what to prepare, how to carry out each step, what evidence or follow-up matters, and when the situation deserves professional or official help.
A ratio is a screening signal, not a legal verdict. Good analysis combines accurate denominators, practical context, statistical testing, and review of the job-related selection procedure.
What You Should Know Before You Start
- The federal Uniform Guidelines use the four-fifths rule as a general screening tool, not an automatic legal conclusion.
- Selection rate equals selected divided by considered for the same stage and period.
- Differences smaller than 20% may still matter, and small samples can make the ratio unstable.
Rules, contracts, platform procedures, and agency practices can change. For legal, financial, tax, veterinary, employment, or safety-sensitive decisions, confirm the current rule that applies to the location and facts before acting.
Quick Preparation Checklist
- Define one employment decision point and one consistent comparison period.
- Check data quality, eligibility rules, group counts, and missing demographic information.
- Preserve the calculation method so another reviewer can reproduce the result.
- Keep the raw numbers and units before rounding.
- Write the formula or comparison rule and test it with a small example.
How to Calculate Adverse Impact: Step by Step
1. Define one employment decision
Analyze screening, interviews, hires, promotions, terminations, or test passage separately.
This is the point where accuracy matters more than speed. Translate the idea into a concrete action: identify who must act, what information they need, and what a successful result would look like. Do not rely on memory when a record, official lookup, written agreement, or dated message is available.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Before moving on, test the step against the real facts of your situation. Ask whether the information is current, whether a deadline applies, and whether another person could verify the same conclusion from the records you have.
2. Choose a consistent period and population
Use the same eligibility and data rules for every group.
Keep the scope narrow. A focused request or calculation is easier to evaluate than a broad accusation or an open-ended demand. Write down the key names, dates, amounts, locations, or decision criteria so that the next step does not depend on assumptions.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. If the answer changes the rights, safety, or money of another person, use a second check. That may mean calling an official number, reading the current procedure, comparing the contract, or asking a qualified professional to review the issue.
3. Count considered and selected
Retain numerator, denominator, and missing demographic data.
Treat this as an evidence-building step, not merely a task to finish. Save the original document or screenshot, record when you obtained it, and note what it proves. A clean record is useful even when the matter is resolved informally because it prevents later disagreements about what happened.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Pause when the evidence conflicts. Do not force an answer by selecting only the facts that support your preferred outcome. Resolve the mismatch, correct the data, or explain the uncertainty before continuing.
4. Calculate each selection rate
Divide selected by considered and keep sufficient decimal precision.
Use plain language and measurable details. Replace words such as “soon,” “a lot,” or “unfair” with a date, amount, frequency, comparison, or requested action whenever possible. This gives the other person or agency something they can actually confirm and respond to.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Keep the response proportionate. Start with the least disruptive option that protects your rights and safety, then escalate only if the problem continues, the deadline requires it, or the available evidence justifies a stronger step.
5. Identify the highest rate
Use that group as the comparison group.
Think about implementation, not only theory. Decide where the information will be stored, who will follow up, what proof will be retained, and when the result will be reviewed. A good plan includes the next checkpoint rather than ending with a vague intention.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Watch for a hidden dependency: an approval, signature, permit, account status, witness, or response from another organization. Identify it early so the process does not stall after the easier work is complete.
6. Calculate the impact ratio
Divide another group’s rate by the highest selection rate.
This is the point where accuracy matters more than speed. Translate the idea into a concrete action: identify who must act, what information they need, and what a successful result would look like. Do not rely on memory when a record, official lookup, written agreement, or dated message is available.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Before moving on, test the step against the real facts of your situation. Ask whether the information is current, whether a deadline applies, and whether another person could verify the same conclusion from the records you have.
7. Apply the four-fifths screen
A ratio below 0.80 generally indicates a need for deeper review.
Keep the scope narrow. A focused request or calculation is easier to evaluate than a broad accusation or an open-ended demand. Write down the key names, dates, amounts, locations, or decision criteria so that the next step does not depend on assumptions.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. If the answer changes the rights, safety, or money of another person, use a second check. That may mean calling an official number, reading the current procedure, comparing the contract, or asking a qualified professional to review the issue.
8. Check sample size and significance
Test whether results are stable, statistically meaningful, and practically important.
Treat this as an evidence-building step, not merely a task to finish. Save the original document or screenshot, record when you obtained it, and note what it proves. A clean record is useful even when the matter is resolved informally because it prevents later disagreements about what happened.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Pause when the evidence conflicts. Do not force an answer by selecting only the facts that support your preferred outcome. Resolve the mismatch, correct the data, or explain the uncertainty before continuing.
9. Investigate and validate the process
Review job relatedness, scoring, discretion, data quality, and less discriminatory alternatives.
Use plain language and measurable details. Replace words such as “soon,” “a lot,” or “unfair” with a date, amount, frequency, comparison, or requested action whenever possible. This gives the other person or agency something they can actually confirm and respond to.
Use consistent denominators and comparison groups. A result that cannot be reproduced from the raw data should not guide an employment decision. Keep the response proportionate. Start with the least disruptive option that protects your rights and safety, then escalate only if the problem continues, the deadline requires it, or the available evidence justifies a stronger step.
How to Build a Stronger Record
Keep the raw counts, formulas, rounding rules, data exclusions, and version history in a controlled file. Organize the material chronologically. Keep original files instead of only edited screenshots, and label each item with the date, source, and what it helps prove.
Separate evidence from argument. First show what happened; then explain why it matters and what outcome you are requesting. This structure helps a company, agency, mediator, lawyer, or court evaluate the matter without searching through emotion or repetition.
Use a short index when there are several documents. An index can connect each important statement to the relevant contract, photograph, message, receipt, official lookup, calculation, or witness.
Practical Example
If Group A selects 60 of 100 applicants, its rate is 60%. If Group B selects 36 of 80, its rate is 45%. The impact ratio is 45% ÷ 60% = 0.75.
The value of the example is its structure: it identifies the facts, the supporting record, and the next requested action. Adapt the names, dates, numbers, and procedure to the real situation rather than copying wording that does not fit.
Common Mistakes to Avoid
- Using the total applicant pool as every denominator: This weakens the record and makes it harder for another person or agency to understand the real issue.
- Mixing job levels or periods: This can create false expectations or cause you to miss the more important deadline, boundary, or verification step.
- Treating 0.80 as proof of illegality: This often turns a solvable problem into a credibility dispute. Use specific facts and preserve original records instead.
- Ignoring unknown demographic data: This may feel faster, but it removes safeguards that exist to protect money, safety, privacy, or legal rights.
- Rounding too early: This makes the outcome depend on assumptions. Replace it with a documented, measurable action.
A Simple Decision Framework
| Question | Recommended focus |
|---|---|
| Is there immediate danger, fraud, or irreversible loss? | Prioritize safety, account protection, emergency reporting, or a temporary legal measure. |
| Does a deadline apply? | Calendar it and complete the required filing, notice, dispute, or response before informal negotiation. |
| Can the issue be resolved directly? | Make one clear, documented request and allow a reasonable response time. |
| Is there an official process or neutral forum? | Use the agency, mediation, appeal, arbitration, or court route with organized evidence. |
| Is the likely benefit worth the cost? | Compare money, time, relationship impact, enforcement risk, and realistic alternatives. |
How to Follow Up Without Losing Momentum
Set a clear follow-up date instead of waiting indefinitely. Refer to the original request, state what has or has not happened, and identify the next proportionate step. Keep the follow-up shorter than the first message and avoid adding unrelated grievances.
If the issue is resolved, save the final confirmation, payment, agreement, corrected record, or closure notice. If it is not resolved, escalate through a channel that has actual authority rather than repeating the same request to someone who cannot act.
When to Get Professional Help
Employment counsel and an industrial-organizational psychologist can assess validation, statistical significance, job relatedness, and less discriminatory alternatives.
Important: This guide provides general educational information, not individualized legal, tax, financial, employment, veterinary, cybersecurity, or safety advice. Rules and deadlines vary by jurisdiction and facts.
Writer’s Opinion
Adverse-impact monitoring should be designed into hiring systems. I would pair the 80% screen with stage-by-stage analysis and validation rather than waiting for a complaint.
The strongest approach is usually the one that protects safety and deadlines while remaining easy for another person to verify. Clarity, documentation, and proportionate escalation add more value than dramatic language or a rushed decision.
Video Guide
[lordai_youtube id="oVzw_hdswYA" title="The Four-Fifths Rule Explained"]
Frequently Asked Questions
What is the four-fifths rule?
A group rate below 80% of the highest group rate may indicate adverse impact.
Does below 0.80 prove discrimination?
No. It triggers deeper factual, statistical, and validity analysis.
What is the denominator?
The number in the group considered at that specific stage.
Should small samples be ignored?
No, but interpret them cautiously and consider additional periods or methods.
Final Checklist
- The exact goal and requested outcome are clear.
- The facts, calculations, or observations are accurate and arranged logically.
- Important evidence is saved in its original form.
- The correct person, business, agency, or court has been identified.
- Any deadline, notice, privacy, or safety requirement has been confirmed.
- The next step is proportionate, lawful, and realistic.
Conclusion
Calculate each group’s selection rate, divide each lower rate by the highest rate, and flag ratios below 0.80 for further review while considering sample size and statistical significance.
Start with the least confrontational step that protects your rights, safety, and evidence. Escalate only when the facts, deadline, repeated behavior, or continuing loss justify it.
