Attention to detail test for hiring: questions, answers, and scoring

What to screen for (and why typo tests aren't enough)

If you're hiring for a role where a wrong digit costs real money, the resume tells you almost nothing. Everyone claims "strong attention to detail." The candidate who transposes an invoice quantity and the one who catches it look identical on paper.

Most attention to detail tests measure one narrow behavior: spotting a typo. That's one slice of the skill. Real work errors come from different failure modes, and a good screen covers all of them:

• Scanning and comparison: noticing differences between two sources (CRM vs invoice, ticket vs SOP).
• Rule compliance: applying thresholds, required fields, and escalation criteria consistently.
• Data validation: catching transposition, omission, and substitution errors in numbers, dates, IDs, and addresses.
• Proofreading with intent: finding mistakes that change meaning, not just aesthetics.
• Procedural adherence: following steps without skipping or inventing assumptions.
• Exception detection: spotting plausible-looking values that are wrong.

The diagnostic edge is scoring not just right vs wrong but the type of each miss: omission, substitution, transposition, rule violation, or formatting mismatch. Two candidates with the same total can be very different hires. The one who misses rule items is a bigger risk in a policy-heavy role than the one who misses a formatting item.

The test: 10 questions with answer key

Give candidates 10 to 12 minutes, no spellcheck or autocorrect tools, and this instruction: "Answer each item based only on the provided information. Don't assume missing details. Choose the best option." If you're using this at volume, randomize item order and rotate a parallel form to reduce sharing.

Q1. Side-by-side comparison (transcription)

A customer's email in the CRM is marisol.chen91@outlook.com. The email in the shipment record is marisol.chen19@outlook.com. What is the issue?

• A) No issue; both are valid
• B) Digits are transposed
• C) Domain is incorrect
• D) Missing character in username

Correct answer: B

Q2. Rule compliance (policy threshold)

Policy: "Refunds over $250 require manager approval. Refunds of $250 or less do not." Which refund requires manager approval?

• A) $250.00
• B) $249.99
• C) $250.01
• D) $250.00 if the customer is new

Correct answer: C

Q3. Proofreading with meaning

Choose the sentence with an error that changes meaning:

• A) Please confirm the recipient's address before shipping.
• B) The patient denied chest pain, shortness of breath, or dizziness.
• C) We can not approve the request without documentation.
• D) The report was reviewed and signed by Dr. Patel.

Correct answer: C. "Cannot" vs "can not" can be style-dependent, but in policy and legal contexts it can introduce ambiguity. In strict documentation, "cannot" is typically required.

Q4. Data validation (date logic)

A form states a start date of 2026-03-18 and an end date of 2026-03-08. What's the best classification?

• A) Acceptable; same month
• B) Likely transposition; end date precedes start date
• C) Acceptable if weekend
• D) Missing time zone

Correct answer: B

Q5. Exception detection (pattern outlier)

All part numbers follow the pattern AA-####-B (two letters AA, hyphen, 4 digits, hyphen, B). Which part number violates the pattern?

• A) AA-2048-B
• B) AA-0284-B
• C) AB-2048-B
• D) AA-7401-BC

Correct answer: C

Q6. Procedural adherence (SOP steps)

SOP excerpt: 1) Verify customer identity (2 identifiers). 2) Confirm the order number. 3) Read back the shipping address. 4) Document the confirmation in the ticket.

A ticket note shows: identity verified (name + DOB), order confirmed, confirmation documented. No address read-back is mentioned. What's the correct finding?

• A) Complete; all critical steps done
• B) Incomplete; missing address read-back
• C) Complete if the address is already on file
• D) Incomplete; missing order number

Correct answer: B

Q7. Numeric accuracy (transposition)

Invoice line shows quantity 1,306. Packing slip shows quantity 1,360. What best describes the discrepancy?

• A) Omission
• B) Substitution
• C) Transposition
• D) Rounding

Correct answer: C

Q8. Formatting rule (standardization)

Rule: "Phone numbers must be stored as +1 (###) ###-####." Which entry is compliant?

• A) (415) 555-0182
• B) +1 415-555-0182
• C) +1 (415) 555-0182
• D) 4155550182

Correct answer: C

Q9. Ticket triage (rule-based classification)

Rule: "Escalate to Tier 2 if (a) payment failed twice OR (b) customer is charged but order status remains 'Pending' for over 30 minutes." Scenario: payment failed once. Customer was charged. Order has been Pending for 42 minutes. What should you do?

• A) Keep in Tier 1; only one payment failure
• B) Escalate to Tier 2
• C) Cancel the order
• D) Ask customer to wait 24 hours

Correct answer: B

Q10. Proofreading (high-impact field)

Choose the best correction for a shipping label line. Current: "1250 W. Harrsion St."

• A) 1250 W. Harrison St.
• B) 1250 W. Harrsion Street
• C) 1250 West Harrison
• D) No change needed

Correct answer: A

Answer key

1. B
2. C
3. C
4. B
5. C
6. B
7. C
8. C
9. B
10. A

Scoring: hiring bands and the miss profile

Raw score

1 point per correct item, 10 points total.

Hiring bands

• 10: near-perfect control. Strong signal. Still confirm with a role-specific work sample before the offer.
• 8 to 9: reliable accuracy. Catches rules and comparisons. Good to advance.
• 5 to 7: solid baseline, but vulnerable to exceptions, thresholds, and "almost right" values. Advance only with a work-sample follow-up.
• 0 to 4: frequent misses. Likely inconsistent rule application and weak verification habits. A risk in any accuracy-critical role.

The miss profile (where the real signal is)

For each missed item, classify the miss:

• Rule violation: Q2, Q6, Q8, Q9
• Transposition: Q1, Q7
• Logic and date validation: Q4
• Pattern and format exception: Q5, Q8
• Proofreading and meaning: Q3, Q10

Match the profile to the role. Missed rule items matter most for policy-heavy work (billing, support escalations). Missed transpositions matter most for data entry and reconciliation. In the interview, show the candidate a miss and ask how they verified their answer. How they respond to their own error is a second signal.

Red flags

• Two or more rule-violation misses for a policy-heavy role.
• A perfect score achieved suspiciously fast on an unproctored screen. Verify with a live work sample.
• Boundary-value misses (Q2-style). These are the errors that trigger refund and compliance incidents.

Confidence notes

• Short tests amplify luck. Don't set rigid cutoffs on 10 items alone.
• Use a two-step process: this screen to prioritize, then a 20 to 30 minute job-relevant work sample for finalists.
• For high-risk workflows, confirm with a realistic artifact review scored on consistent criteria.

Role-based blueprints (extending to a 25-35 item screen)

There's no universal "good score" without job context. When you build a longer version, weight the item mix to the role:

• Data entry and operations admin: 40% comparison and transcription, 30% formatting, 20% rule checks, 10% proofreading. Confirm strong scorers with a data-entry work sample. If this is your hire, see the data entry clerk hiring guide for the full process.
• QA and technical support: 35% rule application, 35% exception detection, 20% comparison, 10% proofreading. Emphasize Q6 and Q9-style items.
• Accounting, AP, and billing: 40% numeric validation, 30% rule thresholds, 20% comparison, 10% documentation quality. Add a reconciliation work sample.
• Healthcare documentation and intake: 40% procedural adherence, 30% identifiers and transcription, 20% proofreading with meaning, 10% formatting. Prioritize omission detection.

Compliance and fairness notes

• Match time pressure to the job. Over-timing an accuracy role turns your screen into a speed test and measures the wrong thing.
• Map the test to real tasks. List the 5 to 7 accuracy-critical tasks in the role and link each to a subskill above. That's your job-relevance documentation.
• Administer identically: same items, same instructions, same time limit for every candidate.
• Accessibility: avoid color-only signals, use readable fonts and sufficient contrast, and provide reasonable accommodations like extra time or assistive tech compatibility when requested.
• Don't infer unrelated traits (intelligence, motivation) from a single short score. It measures accuracy behaviors, nothing more.
• Calibrate internally: run the test with a couple of your current strong performers so you know what scores look like in your context before you set a bar.

Run this screen automatically with Truffle

Grading 10 items is easy. Grading 10 items for 120 applicants, consistently, at 9pm, is how mistakes get hired.

Truffle is a candidate screening platform that combines talent assessments with resume screening and one-way video interviews. Set up this test once and every applicant gets the same items, the same timing, and the same scoring. Truffle ranks candidates against your criteria and shows you why each one scored the way they did, so your shortlist arrives sorted instead of alphabetical. AI surfaces the evidence. You make the call.

The 7-day free trial includes 30 credits, no card required. Start free trial