Skills Assessment Test for Administrative Assistant: Capability Only

A skills assessment test for administrative assistant roles is a bundle of short timed sections rather than one exam. Most employers assemble five: error checking, written communication, organizing and prioritizing, software proficiency in Word, Excel, Outlook or Google Workspace, and situational judgement. Each section is scored on its own.

• Error checking and accuracy. Two records that should match, and a mismatch you have to find under a clock.
• Written communication. A short email drafted to a brief, or a paragraph handed to you to fix.
• Organizing and prioritizing. A crowded morning with more demands in it than hours, put in order.
• Software proficiency. Word, Excel, Outlook or Google Workspace, tested at the depth the job uses.
• Situational judgement. A workplace squeeze with no clean answer, scored against a keyed response.

When an employer says they are sending an administrative assistant skills test, they almost never mean a single test. What arrives is a bundle, assembled from a menu, and the pieces are scored separately even when the report puts one number at the top. Working out which pieces you are facing is most of the preparation.

What is on an administrative assistant test?

The same five blocks recur across almost every platform. Error checking hands you two records that should match and asks whether they do: a purchase order against an invoice, a surname spelled two ways, a date that moved by one digit. Written communication gives you a short brief and asks for the email, or gives you a paragraph and asks you to fix it. Organizing and prioritizing describes a morning with four demands and one hour in it, then asks what you do first. Software proficiency covers Word, Excel, Outlook, or Google Workspace, sometimes as multiple choice about menus and sometimes as a live task inside the real application. The fifth block, increasingly common, is situational judgement: your manager and a client both want the same afternoon, and you choose the least bad response. A clerical skills test is the older and narrower version of the same idea, weighted toward filing, coding, and data checking.

Four administrative assistant skills test questions, with the answers

Here are four items, one from each block you are most likely to meet, written in the style of the test itself. Each gives you the item, the keyed answer, and the reason it is keyed that way, because the reason is the part that transfers to the version you sit.

1. Error checking. Purchase order 4471 reads: 12 units of BX-2290 at $84.50 each, deliver to Suite 300, Elmwood House. Invoice 4471 reads: 12 units of BX-2290 at $84.50 each, deliver to Suite 3000, Elmwood House. Do the two records match? Answer: No. The delivery suite differs, 300 against 3000. Why it is keyed that way: every field a candidate scans hardest, quantity, item code, unit price, is identical, so the planted mismatch sits in the line the eye reads as boilerplate. These items are built to test where your attention drops, not whether you can compare two numbers.
2. Organizing and prioritizing. It is 9:05. Your director needs a board pack printed and bound before an 11:00 meeting. A client is holding on line two to move tomorrow's site visit. Payroll closes at 10:00 and two timesheets are still unapproved. A colleague wants a proposal proofread by end of day. What do you do first, and in what order? Answer: Take the held call and settle the new date, then chase the two timesheets before 10:00, then the board pack, then the proposal. Why it is keyed that way: the key follows hard external deadlines and who is blocked, not who asked loudest. The call costs ninety seconds and a person is waiting on it; payroll is the only item that cannot be done late; the board pack has a genuine ninety-minute runway. The distractor that catches most candidates is starting with the director.
3. Excel task. A sheet carries 480 rows: column A supplier, column B invoice date, column C amount. Show the total spend with one supplier in a single cell. What do you use? Answer: SUMIF, written as =SUMIF(A2:A481,"Elmwood Supplies",C2:C481). SUMIFS if a date range is added. Why it is keyed that way: filtering the column and reading the status bar gets the right number once. The formula gets it again every time a row is added, which is the behaviour the block is paid to detect. Live-task versions score the formula sitting in the cell, not the value you typed into it.
4. Proofreading pass. Mark every error: Please find attached the agenda for Thursdays meeting. The venue has changed to the Elmwood room, and there are two additional item's on the agenda. Their are four attendees confirmed.Answer: Three. Thursdays needs the possessive apostrophe, Thursday's; item's is a plural and takes no apostrophe, items; Their should be There. Why it is keyed that way: proofreading blocks score catch rate against a fixed list of planted errors, and every plant is mechanical, an apostrophe, a plural, a homophone. Rewriting a sentence you simply dislike earns nothing and costs you the clock.

What computer skills do administrative assistants need?

The software block is where candidates prepare hardest in the wrong direction. O*NET's entry for secretaries and administrative assistants lists thirty-one distinct tasks under the occupation, and the software ones are ordinary: entering records into a database, keeping a filing system, running an email system, scheduling and confirming appointments. Employers test to match that. Excel almost always means sorting, filtering, a total or an average, and formatting a table so it prints. It rarely means pivot tables and it hardly ever means macros. Word means styles, tracked changes, and mail merge where the role does mailings. Outlook means calendars with more than one person in them. Our guides to the computer skills test and the Excel test go block by block, and the Office skills exam guide covers the certification version.

How long does an administrative assistant assessment test take?

Published durations for what is nominally the same test run from twenty minutes to an hour and a half, and the sources disagree with each other, so read any single figure you are handed as that vendor's number rather than the standard. The spread has one main driver. A test built from multiple-choice items about software and grammar lands near twenty to forty minutes. A test that makes you do the work, drafting the email into a box, completing a task in a live spreadsheet, typing against a timer, runs sixty to ninety. Ask the recruiter which kind is coming. They will usually say, and the answer changes what is worth rehearsing.

Do admin assistants need good maths?

Admin roles do carry numbers, and candidates worry about them more than the items deserve. What gets tested is arithmetic with a purpose: a percentage off an invoice, a total that should reconcile, an expense column that has to add up, a reading off a small table. There is no algebra and no geometry. If numbers are your weak block, an hour on percentages and one calculator-free total will move your score further than a week of general revision, because the items are not hard, they are fast.

Reviewed by Pigment's Assessment Science Team, who build the Pigment career test and read the technical documentation behind the screening tools described here. Last reviewed 13 September 2026.

Three scoring models are in circulation and they are not interchangeable. A raw percentage tells you how many items you got right and nothing about whether that is good. A percentile or band compares you against a reference group, so one raw score can read as strong or weak depending on who is in that group. A pass or fail recommendation applies a cut-off the employer chose, not one the test author set. The band is the number worth asking about, because it is the one number shaped like the question you brought.

How do you test administrative skills?

Federal guidance on employment testing sorts the instruments into categories that map onto an admin screen cleanly: cognitive tests, covering reasoning, memory, perceptual speed and accuracy, arithmetic and reading comprehension; sample job tasks, which include work samples and simulations; and personality or integrity tests. An administrative assistant screen is usually the first two stitched together. Each reads something different. A cognitive block reads speed and accuracy under pressure. A work sample reads whether you can produce the artifact at all. Neither one watches you produce it every Thursday for nine months, which is the part the job is made of. Our guide to hiring aptitude tests walks the employer side of that choice.

No pass mark on most of them, so the ranking is what turns you down

Most employers rank. They open a req, collect thirty results, and read down from the top, which makes a good score a position rather than an absolute. One prep platform issues a certificate to the top quarter of its own test-takers, which is a ranking wearing the clothes of an achievement. If you are told a cut-off exists, ask what it is and who set it. Title VII restricts an employer from using different cut-off scores for different protected groups, so a cut-off is a documented decision somebody owns, not a house habit.

Sixty-eight means nothing until you know what the benchmark group was

A hiring manager opens the report and sees a composite plus a breakdown by block. The composite is the number quoted in the meeting. The breakdown is the number that decides anything, because a 68 built from one weak software block and strong everything else is a fixable candidate, and a 68 built from thin accuracy throughout is not. The field that builds these instruments is clear that a score without a reference group is not yet a result. And the duty to establish that a test is any good sits with the employer rather than the vendor: the EEOC states that vendor documentation may help, and that the employer remains responsible for its tests being valid for the job they are used on. Almost nobody on either side of the desk asks for that evidence.

How to tell whether an administrative assistant skills test is any good

Three properties decide whether a test deserves the weight being put on it, and all three can be asked about in one email. Reliability is consistency: would the same person, sitting the same test on a different Tuesday, score roughly the same? A test that swings ten points on mood is measuring the mood. Validity is whether the score relates to the work: has anyone checked that people who score well here go on to do this job well, in this kind of role? Correlations with other tests are often offered instead, which is a weaker claim wearing the same word. The benchmark group is who your percentile is measured against, and to mean anything it has to be a sample of people doing the job you are hiring for or applying to, described by size, date and makeup, rather than the vendor's whole database of everyone who ever clicked through. What to ask for is short: the technical manual, the reliability figure for the specific blocks in your screen, the validity study and the population it ran on, and the benchmark group's size and composition. The federal Uniform Guidelines on Employee Selection Procedures set out what that evidence has to look like. A vendor who cannot produce it has answered the question.

Read the report in both directions, because a high number and a low one are each less dramatic than they look on the page.

A high score means the mechanics are not your constraint

If you cleared accuracy, writing, and software comfortably, you have learned something true and narrow: the daily mechanics of the job will not be what defeats you. That is worth knowing before an interview, because it lets you stop rehearsing the part you already hold and start preparing for the part nobody scored, which is the conversation about the team, the calendar, and who gets to interrupt you.

What strengths should admin assistants show?

Strip the marketing off any administrative assistant screen and the same five behaviors are being paid for. Vigilance, which is noticing the thing that does not match. Detail orientation. Structure creation, which is building the system rather than working inside somebody else's. Completion, which is closing the loop instead of opening the next one. Coordination, which is holding four people's time in your head at once. Those are the strengths to bring to an interview, and each one lands better as an example than as an adjective. A dedicated attention to detail test reads the first two on their own.

A weak block is usually a tool you have not used

Low software scores are the most common single failure and the most fixable one. Somebody who has run a department on Google Sheets for four years will score badly on an Excel-specific block and be entirely capable of the job. The same goes for a proofreading block for anyone whose last role had a copy editor beside them. Two weeks of deliberate practice moves it. A thin accuracy score under time pressure is a different signal and worth taking more seriously, because that one tends to repeat.

The situational judgement block scores the answer an experienced colleague would give

Situational judgement items have no right answer in the ordinary sense. They have a keyed answer, built from what experienced people in the role chose when the item was written. Answer as the person who does this job well, not as the person you are on a bad Tuesday. That is also why a strong situational judgement score says more about your read of the role than about how you will feel doing it in year two.

Four kinds of people arrive at this search. The same instrument is right for three of them and wrong for the fourth.

A test booked for next week: rehearse the shape of it

If a date is already in your calendar, ask the recruiter which blocks it contains and which application the software section runs inside. Then spend your time on the shape of the test rather than on general revision: one timed run at error checking, one email drafted to a brief in under ten minutes, one hour in whichever spreadsheet they named. Our worked administrative officer questions are the closest rehearsal we publish for the reasoning blocks.

How do I prepare for an administrative interview?

The test and the interview reward different preparation, and most candidates over-prepare one of them. For the interview, bring three worked examples instead of adjectives: a time you caught something before it went out, a week where four things landed at once and you sequenced them, and a system you built that somebody else still uses. Read the posting for the software it names and be ready to say how deep you go in it. Then have one question ready about how the team handles interruptions, because that answer tells you more about the seat than your score will.

No admin experience is less of a handicap on the test than on the CV

A skills test is the most winnable part of an application for a career changer, because it marks the task in front of you and has no view of the CV behind it. Nobody is checking whether the spreadsheet you filtered sat in an office. Most of the test is not about experience at all. The exception is the situational judgement block, where the keyed answers reflect how people who have held the role behave, and that is the piece worth reading about before you sit down.

Already in the seat and worn out: this is the wrong instrument

If you are three years into an admin role and weighing whether to stay, a skills test will hand back a good score. You already knew you could do the job. Scoring well on the thing draining you settles nothing, and this is the most common way the search ends in a worse decision. The question underneath is whether the conditions of the work suit you, and a capability test has no access to it.

Hiring managers should read the block scores, not the composite

If you are choosing a screen, the useful output is the breakdown, not the headline. A composite hides the difference between a fixable software gap and a persistent accuracy problem. Two things are worth doing before you buy. Ask the vendor for reliability and validity evidence for this specific role, and check what the benchmark group is, because a percentile against the vendor's whole database is not a percentile against your applicant pool.

Weight the blocks against the req you actually opened, not against the vendor's default. A role that lives in a shared calendar and a shared inbox is mostly coordination and written clarity, so a strong Excel block on that hire is a bonus you should not let outrank the writing sample. A role sitting on invoice reconciliation inverts it: accuracy and the spreadsheet block carry the decision and an adequate email is fine. Write the weighting down before the first result lands, because once you are looking at named candidates the composite quietly becomes the weighting, and the composite was built by somebody who has never seen your req.

When a work sample beats a multiple-choice screen

A multiple-choice block is cheap, fast and consistent, which is why it belongs at the top of a funnel holding ninety applicants and one afternoon. It buys you a rank, not a demonstration. A work sample is the opposite trade: slower to set and slower to mark, and the only version of the test that shows you the artifact the job will ask for. Use the multiple-choice screen to cut the pile, then hand the last four or five people a task from the desk itself, a calendar clash to resolve or a half-page to tidy, timed the way the day will time them. Where the role is mostly production, drafting, formatting, reconciling, the work sample should outweigh everything else combined, because it is the only block that has watched the person do the work.

Start from what you are trying to settle. The instrument follows from that. Sit the wrong one and you walk out with a solid read on your error-checking speed while the thing you were actually weighing, whether this seat is worth taking, never came up.

If an employer named the test, practice that test

Generic practice helps less than people expect. The formats differ enough that rehearsing an error-checking grid will not prepare you for a live Excel task, and drilling Excel will not prepare you for a proofreading pass against a timer. Ask for the name of the platform and the blocks inside it. If the recruiter will not say, prepare the two that turn up almost everywhere: error checking, and one office application.

Free practice packs rehearse the format, and that is most of the benefit

Free administrative assistant practice tests are worth taking, for one narrow and genuine reason: the first time you meet a timed error-checking grid should not be the time it counts. What they cannot hand you is a score that means anything, because no benchmark group sits behind them, and they cannot tell you how a particular employer weights the blocks. Take two or three, note which block slowed you down, then stop. A fuller account of what free tools do and do not deliver sits in our free skills test guide, and the broader skills assessment guide covers the other formats.

Deciding whether the seat fits: a capability test is the wrong tool at any quality

This is the case where a better test makes things worse. A more rigorous screen produces a more confident answer to a question you were not asking, and a strong result then reads as permission. If what you are weighing is whether to take the offer, or whether to stay another year, no capability instrument reaches it. The better the work suits someone, the less they want out, and that relationship is a correlation of -.46 across the research on how people fit their work, a result that dates to Kristof-Brown and colleagues in 2005. An admin role tends to be settled by the calendar, the interruptions, and who owns your afternoon, long after the score has been forgotten.

What a career read adds, and what it refuses to do

Pigment's Career Self-Discovery Assessment is forced-choice, and it reports to nobody but you. It returns a 36-page read: the strengths you carry, the way you take information in and reach a decision, the rare trait combinations that make you unusual, and named roles with the shape of the week each one implies. Outside researchers have not yet checked those role matches against what people went on to do, so it cannot tell you the roles it names will still hold a year into the seat. It is not a selection instrument either, so no employer should ever be scoring a candidate with it. For the wider field of instruments and what each is for, our career test guide maps them.

The score answers whether you can. Nothing on it answers whether you should

The test in front of you is a fair instrument for a fair question, and if it is booked for Tuesday you should prepare for it properly. Keep the two readings apart afterwards. A strong score means the mechanics of the job will not defeat you. Whether the work will still hold you three years from now is a different reading, it was never on the report, and it is yours to run on your own account.