Back to the on-screen lesson ·
Hire on evidence that bears on the written role — a paid work sample scored against the standard and references about specific behavior — with the same task and scoring for every candidate, and irrelevant characteristics kept out.
Paper packet. Every task here also exists on screen, where it is checked automatically; answers written on paper are not assessed by Nydus. When you are back at a device, enter your answers there.
You will choose hiring evidence that bears on the role, sort evidence from impressions and from characteristics that must not be used, write reference questions about behavior, put a fair hiring sequence in order and total candidates' scores.
A written role says what the person will deliver and how it is measured. That is also what hiring should test: not whether the owner liked the candidate, but whether they can do this work to this standard.
| Term | What it means |
|---|---|
| Hiring evidence | Information that bears on the role's outcomes. |
| Work sample | A short, paid piece of the role's real work, scored. |
| Structured reference | Questions to a former employer about specific behavior. |
| Scoring sheet | The criteria, fixed before any candidate is seen. |
| Impression | A feeling about a person. |
| Protected characteristic | A trait the law forbids using in hiring decisions. |
The most common hiring method in small businesses is a chat. It measures how comfortable the owner felt, which depends on confidence, charm and how much the candidate resembles the owner, none of which is plating lunches to a standard or servicing a bike safely.
The strongest evidence is a work sample: the candidate does a short, paid piece of the real work. Neighborhood Kitchen asks each candidate to plate lunches from the written standard for 45 minutes, scored for accuracy and time. Northside Repairs asks for a timed service of a test bike, scored for the safety checks, the standard of the work and the time taken. The scoring sheet is written before anyone arrives, and every candidate gets the same task and the same scoring; otherwise the criteria drift toward whoever was liked.
References help if they ask about behavior, not opinion: 'how often did their cleans need redoing?' rather than 'are they nice?'.
Another way: table
Evidence, ranked by how much it tells about the role.
| Evidence | What it shows |
|---|---|
| Paid work sample, scored | The actual work, to the actual standard |
| Reference about specific behavior | What they did in similar work |
| Past jobs listed | Where they have been, not how well |
| A pleasant chat | How the owner felt |
Another way: steps
Some characteristics have nothing to do with the work and must not enter the decision: a candidate's age, whether they plan to have children, their religion, race or national origin, a disability that does not prevent the work with reasonable accommodation, and others. Using them is unfair, discards good candidates, and is unlawful. Federal laws such as Title VII of the Civil Rights Act, the Age Discrimination in Employment Act and the Americans with Disabilities Act set the baseline, and many states add more protected characteristics and cover smaller businesses. Employment rules — pay, leave, contracts, working time, safety, ending employment — are set where the business operates and change over time. Look up the current rule from an official source for your own jurisdiction, and note the source and the date you checked it.
Start from the role. Its outcome, responsibilities and measure say what the evidence should test.
Choose the work sample. A short piece of the real work, an hour or less where possible, that every candidate can do on the same equipment. Pay for it.
Write the scoring sheet. Three or four criteria from the standard work, each scored the same way, such as out of 5. Decide any weights now.
Run the samples. Same task, same instructions, same time limit, same scorer where possible.
Ask references. For the leading candidates, questions about specific behavior in similar work.
Decide and record. The highest total, unless a reference raises a concern about the role's work; write down the reason either way.
Check the process. Could every question asked be traced to the role? Did every candidate get the same sample and scoring? Six months after the hire, compare the person's measure with their sample score: if the two disagree badly, the sample is testing the wrong thing.
Scoring a work sample is allowed because it shows the thing the role needs done. A person who plates accurately in the trial is likely to plate accurately on the job; a person who chats well is not shown to do anything the role needs.
Fixing the scoring sheet in advance is needed because people judge more kindly those they liked on first sight. Criteria chosen afterwards tend to favor them; criteria chosen before cannot.
Using the same sample for everyone is allowed, and fair, because only then do the scores compare. A candidate given an easier bathroom to clean has not shown more skill.
Recording the reasons is needed because a decision that cannot be explained cannot be checked, by the owner or by anyone else.
A good sample is short, real and safe. Short, because candidates have other commitments and the owner has a business to run. Real, because a puzzle or a quiz tests something other than the work. Safe, because a candidate who does not yet know the shop should not handle dangerous equipment unsupervised.
Pick the part of the role that matters most and varies most between people: plating for the kitchen, a bathroom clean for the cleaner, a standard service for the mechanic. Give the written standard in advance, so candidates are judged on doing the work, not on guessing what the owner wants. Pay at least the minimum wage that applies, for every hour of the sample; whether trial work must be paid is set by wage law, and paying is both fair and the safe course.
Interviews still have a place, as long as they ask the same questions of everyone and the questions come from the role. Ask about specific past situations: 'tell me about a time you found a mistake in your own work after it was finished; what did you do?' Score the answers on the sheet, like the sample.
Avoid questions that invite protected information, even casually: family plans, age, health, religion, where someone's accent comes from. A friendly question can still collect information the decision must not use, and it can make a fair decision look unfair. If a candidate raises such a topic themselves, do not pursue it.
Hiring evidence is itself a process that can be improved. Keep the scoring sheets, and after six months compare each hire's sample score with how they are doing on the role's measure. If high scorers do well and low scorers would have struggled, the sample is testing the right thing. If not, change it: perhaps it was too short to show consistency, or it tested speed when the role needed care.
The same sheets are the honest answer to a candidate who asks why they were not chosen: 'your sample scored 9 out of 15, and the person hired scored 13; the main difference was the time taken'. Feedback like that is rare, valued, and protects the business by showing the decision rested on the work.
Small businesses often hire friends of friends, former customers or family. That can work well, and it carries a particular risk: the owner already has an impression, and the impression fills in for evidence. Use the same work sample and the same scoring for a candidate the owner knows as for a stranger. It protects the business from a poor fit, and it protects the relationship, because a decision made on the work is easier for everyone to accept than one made on friendship.
The same goes for referrals from current staff. They are often the best source of candidates, because staff know the work, but the referred candidate still does the sample. A referral is a reason to shortlist someone, not a reason to skip the evidence. Thank the member of staff either way, so referrals keep coming even when a referred candidate is not the one hired in the end, and say why.
Federal rules require employers to keep hiring records, such as applications and notes, for a set period, and to complete an I-9 form for each new hire to verify eligibility to work in the United States. Many states and some cities add their own rules, such as limits on asking about criminal records or salary history before an offer.
Look up the current rules for the business's state and city from official sources, such as the U.S. Equal Employment Opportunity Commission and the state labor department, before advertising, and note the source and the date. A scoring sheet kept with the application makes following them much easier.
An independent auto repair shop in Denver had hired three technicians in two years on the strength of interviews, and two had to be let go for poor work within months. The owner changed the process. Each shortlisted candidate is now paid for a two-hour work sample on a shop car with known faults: diagnose two problems, write up the findings, and complete a brake job to the shop's written standard.
The scoring sheet, written before advertising, covers correct diagnosis, the write-up a customer could understand, the brake job against the standard, and safe working, each out of 5. References are asked how often the candidate's work came back and how they handled a job they could not finish on time.
The next two hires scored 17 and 18 out of 20. Both were still at the shop two years later, with comebacks below the shop average. The owner also found that two candidates who had interviewed well in the past would have scored poorly on the sample.
The U.S. Equal Employment Opportunity Commission runs a small-business resource center that explains which federal anti-discrimination laws apply at which business sizes, what may be asked in hiring, and how to keep records. It is a good first official source, alongside the state's own fair-employment agency.
A good chat shows a good hire. It shows a comfortable conversation.
Confidence is competence. A confident candidate may do the work worse.
Decide the criteria after meeting people. Then the criteria follow whoever was liked.
Any question is fair in a small business. Some characteristics must not be used, wherever the business is.
A trial shift can be unpaid. Pay for work samples; wage law decides, and fairness agrees.
Start from the role.
$\text{every clean passes the walk-through}$
The outcome to test.
Choose the sample.
$\text{a paid hour cleaning a bathroom}$
Real work, short.
Write the criteria.
$\text{standard met; checklist followed; time}$
Out of 5 each.
Fix them in advance.
$\text{before any candidate is seen}$
So they cannot drift.
Run it the same way.
$\text{same bathroom, same products, same scorer}$
Comparable scores.
Start from the role.
$\text{bikes to standard, ready on time}$
The outcome.
Choose the sample.
$\text{a timed service of a test bike}$
Paid.
Write the criteria.
$\text{safety checks; standard of work; time}$
Out of 5 each.
Score candidate A.
$5 + 4 + 3 = 12$
Out of 15.
Score candidate B.
$3 + 4 + 5 = 12$
The same total.
Break the tie on the role.
$\text{safety weighs most: A}$
A rule written in advance.
Write the role.
$\text{packing, trimming, cleaning; 14 hours}$
From the delegation list.
Write the scoring sheet.
$\text{accuracy; care with fragile pieces; time}$
Before advertising.
Run the same sample.
$\text{three real orders packed, paid}$
For three candidates.
Total the scores.
$12; \ 9; \ 13$
Out of 15.
Ask references about behavior.
$\text{how they handled breakages}$
For the top two.
Decide and record.
$\text{hire the 13; sheets filed with reasons}$
The record of why.
Check after six months.
$\text{orders out on time: 98 percent}$
The sample predicted well.
Choose the sample.
$\text{a paid 45 minutes plating lunches}$
Real work.
Write the criteria.
$\text{accuracy (counted twice); speed; clean station}$
Fixed in advance.
Write a reference question.
Neighborhood Kitchen is hiring a lunch assistant whose outcome is plates to the standard, on the pass quickly. Which evidence bears most directly on that?
Complete the worked solution: Bright Home Cleaning scored two candidates' paid trial cleans out of $5$ each for standard met at the walk-through, checklist followed, and time. Candidate A scored $4$, $5$ and $4$; candidate B scored $1$, $1$ and $1$. Find each total and how far apart they are.
Find candidate A's total.
$(\text{standard}) + (\text{checklist}) + (\text{time}) =$ x
The same three criteria.
Find candidate B's total.
$(\text{standard}) + (\text{checklist}) + (\text{time}) =$ y
Scored the same way.
Find the gap.
$(\text{A's total}) - (\text{B's total}) =$ g
How far apart on the evidence.
Say what to keep.
$\text{the sheets, with the reasons}$
The record of why.
Dev is hiring a second mechanic. Sort each thing he might consider.
| Relevant evidence | An impression | Must not be used | |
|---|---|---|---|
| A timed service of a test bike, scored against the standard | |||
| He reminds me of myself at that age | |||
| Whether the candidate plans to have children | |||
| A former employer on how often their repairs came back |
Bright Home Cleaning's draft questions for a candidate's references. Mark every question about specific, relevant behavior.
This task has no paper form; do it on a device.
Maya is hiring a studio assistant. Put her hiring steps in order.
Number the steps in order (write the number in the box):
Neighborhood Kitchen's scoring sheet counts accuracy to the plating standard twice, because a wrong plate is worse than a slow one. A candidate scored $2$ for accuracy, $4$ for speed and $4$ for keeping the station clean, each out of $5$. What is the candidate's weighted score, out of $20$? Fill in each figure on the sheet.
| Amount | |
|---|---|
| Accuracy score, doubled | |
| Weighted score |
A landscaping company in Kansas City pays each shortlisted candidate for a $2$-hour work sample: planting and mulching a bed to the written standard. It shortlists $5$ candidates and pays $19$ dollars an hour. What does the round of work samples cost, in dollars?
Answer:
A florist scored two candidates' trial bouquets out of $5$ on accuracy to the order, finish and time. Candidate A scored $3$, $4$ and $2$; candidate B scored $3$, $3$ and $5$. On this evidence, who did the role's work better?
Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.
Northside Repairs scored two candidates' timed services out of $5$ each for safety checks, standard of work and time. Candidate A: $5$, $2$, $4$. Candidate B: $4$, $2$, $4$. Fill in each candidate's total.
| Total out of 15 | |
|---|---|
| Candidate A | |
| Candidate B |
You can hire on what the candidate can do rather than how the conversation felt. Tell someone why the scoring sheet is written before anyone is interviewed. Next: training the person you hired.
19. Your turn: Neighborhood Kitchen's lunch assistant, step 3
$\text{how did they handle a rush?}$
About behavior.