The Real Cost of Cognitive Testing Platforms

0 Quick Answer: What It Costs to Test Other People

Administering a mainstream cognitive battery to another person costs roughly $20 to $35 per administration in software and materials, and roughly $90 to $100 per administration in professional time, so the platform invoice is usually a quarter of the true cost. Those two halves are priced in completely different places, which is why almost nobody publishes a combined number.

The software half is knowable. Pearson prints a public sheet for Q-interactive, its iPad based administration system, and the Pearson store prints WAIS-5 kit prices to the cent. Checked on September 3, 2026, a Q-interactive Standard License costs $310 per user per year for a practice of one to four examiners, and a comprehensive subtest costs $1.95 each time it is administered. A ten subtest primary battery therefore costs $19.50 in usage.

The labor half is not printed anywhere, because it is your payroll rather than a vendor's price. It has to be modelled from wage data and from the administration time the publisher states. That is the honest complication in this whole category: every directory page ranking for this query publishes a feature grid and no arithmetic, because arithmetic requires committing to an assumption in public.

This page commits. Every vendor figure below is quoted from a source you can open, with the date it was checked. Every modelled figure is labelled as modelled and shows its inputs. Where a vendor publishes nothing, the cell is left empty rather than filled with a guess. What the money buys, as opposed to what it costs, is a separate question answered at what makes an instrument professional grade.

1 Two Cost Questions, and the Five Lines in the Answer

There are two entirely different questions hiding behind the phrase cognitive testing cost, and mixing them produces numbers wrong by a factor of thirty. The first is what an individual pays to be tested. The second is what an organization pays to test other people. This page answers only the second.

The first is already answered at the consumer price ladder that runs from free screeners past six thousand dollars. If you are the person being assessed, that is your page, because nothing below is what a clinic charges you. It is what the clinic spends. The gap between the two is not a scandal: a practice billing four thousand dollars is not marking up a twenty dollar subtest bundle, it is billing many doctoral hours across the interview, administration and feedback stages a supervised assessment actually contains, plus supervision, premises and cancelled appointments.

Five lines make up the answer, and vendors publish at most two of them. The seat or license cost is annual, per examiner, and independent of volume, which punishes low volume users hardest. The per administration usage fee is charged per subtest on Q-interactive, per report on a report based platform, and per participant session on some research systems. The kit and its consumables split into a one time capital cost that amortizes and a record form that recurs for every single person tested.

The fourth line is examiner time, the hour in the room plus rapport, setup and breaks. The fifth is scoring, interpretation and reporting time. Digital administration removes hand scoring arithmetic and removes none of the interpretation, because somebody still has to read the profile and write it down. Pricing that hour honestly means knowing what the profile contains, which means the six CHC domains a broad battery has to sample, what an individual subtest type is measuring, and the shape of the multi level report a full profile produces.

Hardware is a sixth line for tablet systems, and this page leaves it blank on purpose. Q-interactive runs on two compatible iPads per examiner, and Apple's list price, institutional discounts and refurbished pricing all differ enough that no single figure would be honest. Whether the instrument itself is the right one, rather than the cheapest, belongs with supervised administration set against online administration.

2 What Q-interactive Actually Costs, From the Published Sheet

Pearson publishes a complete Q-interactive price list as a public PDF, and it is the single most useful document in this category. The file is titled the Q-interactive Product Codes and Pricing Sheet, its embedded document properties name it Q-interactive 2026 General Pricing, it carries the production code C15251 EL 11/25, and it was retrieved from Pearson's Q-interactive pricing sheet on September 3, 2026.

The structure is an annual license per user plus a usage charge per subtest administered. Pearson's own Q-interactive product page states that the license runs for one year from the date of purchase, that usage is measured in subtests, and that buyers choose between pay as you go invoicing and prepaying an estimated annual volume for a tiered rate.

The Standard License matters most for adult cognitive work, because it is the tier that includes WAIS-5, WAIS-IV, WISC-V, WMS-5, NEPSY-II, D-KEFS and RBANS Update. It is priced in three bands: $310 per user per year for one to four users, $285 for five to twenty four, and $260 for twenty five or more. An Academic License limited to WIAT-4, KTEA-3, WRAT5, BBCS-4:R and BSRA-4 runs $115, $95 and $75 across its own bands. A Speech and Language License runs $210, $185 and $160.

Usage is bucketed by task size rather than by test name, which is the detail most summaries get wrong. Academic subtests are $0.98 per administration. Comprehensive subtests, the bucket holding WAIS-5, WISC-V, WMS-5 and D-KEFS, are $1.95, so a short working memory subtest such as Digit Span costs exactly the same as a long one. Large single tests including CVLT-3, RBANS Update, PPVT-5 and EVT-3 are $5.85 each. Those are the monthly billed rates.

Prepaying changes the rate. Under the site license product code the sheet lists $1.95 per subtest for up to 750 subtests, $1.70 for 751 to 5,000, and $1.45 for 5,001 or more. The sheet also states that one set of starter kits per license comes free with all new orders, WPPSI-IV excepted, so the paper materials the system still needs are not an extra first year cost for most buyers. None of this applies to the online WAIS style administrations that are not the publisher's own product, which are a different market with different economics, and the index structure all these subtests feed is set out at the WAIS-5 battery and what its indices contain.

3 What the WAIS-5 Paper Kit Actually Costs

The Pearson store prints WAIS-5 prices to the cent, and the complete print kit is $1,784.00. Every figure in this section comes from the Pearson WAIS-5 product page, read on September 3, 2026, in US dollars.

The complete kit with a soft case is $1,889.00. The two manuals are sold separately at $340.60 each: the Administration and Scoring Manual, and the Technical and Interpretive Manual. Record forms are $193.40 for twenty five, which is $7.74 per person tested. Response booklets are $128.10 for twenty five, which is $5.12 per person, and they exist because the Coding subtest requires a printed response sheet the examinee writes on and Symbol Search consumes the same booklet. Those two consumables together are $12.86 of paper for every administration, and they recur forever.

Scoring is priced separately again. A WAIS-5 Q-global Scoring Subscription for one year is $60.00. A single Q-global Score Report is $3.00, and a report with demographically referenced scores is $4.70. The store lists both the annual subscription and the per report charge without stating whether holding the subscription removes the per report fee. That is a question for Pearson before you build a budget on it, and this page will not assume an answer.

The instrument's own structure drives the arithmetic. Pearson's page states an age range of 16:0 to 90:11, publication in 2024, roughly 45 minutes of administration for the FSIQ and roughly 60 minutes for the ten primary index subtests, and it states that the Full Scale composite now rests on seven subtests rather than the ten used in the previous edition. So on Q-interactive an FSIQ costs 7 times $1.95, which is $13.65, while the full primary index battery costs $19.50.

That reduction is a genuine cost change disguised as a psychometric one, and the substantive account of it sits at what changed between WAIS-IV and WAIS-5 rather than in any price list. Neither of those pages is a rate card, and this one is not an instrument review.

4 Every Published Price, With Its Source and Date

Here is the complete set of figures this page verified against a primary source, in one table. Anything a vendor declines to publish is marked as such rather than estimated. All entries were checked on September 3, 2026.

Two things stand out. Pearson is the only vendor here publishing a full rate card, which makes it the only one you can budget against without a sales call. And the spread between the cheapest and dearest per administration figure is enormous, which means a per administration comparison is close to meaningless until you fix what an administration includes.

That is the same failure that makes consumer comparisons useless, described at free screeners set against validated instruments, and it is why an evidence first ranking such as how the paid online batteries compare on published evidence is more useful than a price ranking. The wider family this all sits inside is mapped at what counts as a cognitive ability test.

5 Paper Versus Digital: The Five Year Arithmetic

Over five years at twenty administrations a year, the WAIS-5 paper route and the Q-interactive route land within about a dollar and thirty cents per case of each other, so the choice between them is not a money decision. The usual assumption, that digital is obviously cheaper or obviously dearer, survives contact with the published numbers in neither direction.

The paper route over one hundred administrations: the complete print kit at $1,784.00, plus one record form and one response booklet per person at $7.74 and $5.12, which is $12.86 times one hundred, or $1,286.00, plus five annual Q-global scoring subscriptions at $60.00, which is $300.00. The five year total is $3,370.00, or $33.70 per administration.

The Q-interactive route over the same one hundred administrations: five annual Standard Licenses at $310 for a single examiner, which is $1,550.00, plus ten comprehensive subtests per administration at $19.50, which is $1,950.00. The five year total is $3,500.00, or $35.00 per administration.

The shapes differ even though the totals do not. Paper is front loaded, so $1,784 leaves the account in month one and the marginal case is then cheap. Q-interactive is flat, so nothing large leaves the account but the seat fee is charged whether you test twenty people or none. A practice with irregular volume is exposed differently by each, and the same split applies to any full kit instrument, including the Stanford-Binet 5 with its own kit and record form structure.

Battery length changes the comparison more than format does. A single domain instrument such as Raven's 2, which samples inductive reasoning alone, runs a fraction of the time and therefore a fraction of the cost, because it asks only for matrix items of the kind that measure Gf and skips five other domains. That is a scope decision priced as a format decision, and it is worth separating the two before signing anything.

What actually separates paper from digital is time rather than money. Hand scoring a battery and checking the arithmetic is examiner time that digital administration removes. Since the next section shows time is roughly three quarters of the total, a route saving twenty minutes per case beats a route saving one dollar per case by a wide margin.

6 The Examiner Hour Is the Whole Answer

In every scenario this page models, professional time is between 73 and 83 percent of the total cost, which means the platform decision is not the cost decision. Pricing that hour requires a wage figure and a time figure, and both are available from primary sources.

For the wage, the Bureau of Labor Statistics Occupational Outlook Handbook entry for psychologists reports May 2025 median annual wages of $100,580 for clinical and counseling psychologists, $95,990 for school psychologists, $193,950 for industrial and organizational psychologists, and $99,110 for psychologists overall, which the same page states as $47.65 per hour. Dividing the clinical and counseling median by a 2,080 hour year gives $48.36 per hour; the school psychologist median gives $46.15.

Those are wages, not the cost of employing somebody. Payroll taxes, benefits, supervision, malpractice insurance, premises and unbilled administrative time all sit on top, and a fully loaded rate is commonly a substantial multiple of the wage. This page uses the bare wage deliberately, which makes every scenario below a floor rather than a forecast. Your real number is higher.

For the time, Pearson states roughly 45 minutes to administer the subtests producing the FSIQ and roughly 60 minutes for the ten primary index subtests. Those are floors too, because they exclude rapport, instructions and breaks, and because timed subtests set the pace of a session rather than the examiner. The scenarios below use two hours of professional time per case: one hour of administration, one hour covering setup, scoring review, interpretation and a short report. That second hour is an explicit modelling assumption, not a published figure, and it is the first input you should replace with your own timesheet data. It is unlikely to be a generous one: the account of where a testing appointment's minutes actually go tracks that same 45 minute instrument becoming a two to three hour session once intake, rapport, verbatim instructions, hand scoring of open responses and breaks are counted, with the report written after all of it.

One consequence follows immediately. A platform that saves fifteen minutes per case is worth about twelve dollars per case at these wages, which exceeds the entire ten subtest usage fee. Evaluate platforms on the minutes they remove, not the cents they charge. The stopwatch heavy part of the battery is the smallest and most fiddly, described at the processing speed tasks that carry most of the timing work, and the reason composites are reported the way they are, which drives report length, is set out at why composites use a standard deviation of 15.

7 Scenario One: One Clinician, Twenty Adults a Year

A solo practitioner administering twenty adult batteries a year spends $700 on the platform and $1,934 on their own time, for a modelled floor of $131.72 per administration. This is where the seat fee bites hardest, because a fixed $310 spreads across only twenty cases.

The platform side: one Q-interactive Standard License in the one to four user band, $310.00 for the year. Usage at ten comprehensive subtests per case, $19.50 per administration, times twenty, which is $390.00. Starter kits free with the new order. Platform subtotal $700.00, which is $35.00 per administration.

The labor side: two hours per case at $48.36, which is $96.72 per administration, times twenty, which is $1,934.40. Total $2,634.40, or $131.72 per administration, of which 73.4 percent is time.

Cutting to the seven FSIQ subtests drops the platform side to $13.65 per case and the annual total to $583.00, saving $117.00 across the year. It also removes the index level detail that makes a profile worth reading, so it saves about six dollars a case and costs the ability to say anything about the pattern. That is usually a bad trade, and it is the kind of trade a cost model surfaces and a feature grid hides. What the broader administration buys instead is described at what a comprehensive battery has to cover and a Full Scale administration set against a single index.

The genuinely large lever is volume. At twenty cases the seat fee is $15.50 per administration; at one hundred cases it is $3.10. A practitioner who tests occasionally pays a structural premium no negotiation removes, which is a real argument for referring out rather than holding a license, and a real argument for the classroom or experiential licenses where a training context qualifies.

The clinical value of the case is created after the scoring finishes, in the interpretation that follows how an index profile is read once the numbers exist. That hour is the product. Everything priced above is the cost of getting to it.

8 Scenario Two: A Research Team, Two Hundred Participants Once

Two hundred supervised WAIS-5 administrations cost a research team $4,020 in platform fees and $19,344 in researcher time, for a modelled $116.82 per participant, and that is before anyone pays the participants. Volume changes which levers matter and makes the prepay band worth using.

Platform side: two Q-interactive seats at $310, which is $620.00. Two hundred participants at ten comprehensive subtests each is 2,000 subtests, which falls in the 751 to 5,000 prepay band at $1.70, giving $3,400.00. Monthly billing at $1.95 would cost $3,900.00, so prepaying saves $500.00 on this single study. Platform subtotal $4,020.00, or $20.10 per participant.

Labor side: two hours per participant at $48.36 is $19,344.00 across the study. Total $23,364.00, or $116.82 per participant, of which 82.8 percent is time.

Participant payment is a third line clinical models omit and research budgets cannot. Prolific's published pricing states a minimum participant rate of $8.00 per hour, a recommended rate of $12.00 per hour, and a platform fee of 33.3 percent for academic and non profit customers or 42.8 percent for corporate customers. Two hundred one hour sessions at the recommended rate is $2,400.00 in rewards plus $799.20 in academic fee, which is $3,199.20, or $16.00 per participant.

The comparison is real and the caveat is larger than the saving. Removing the examiner hour halves the total and also removes observation, environmental control and identity assurance, which is exactly the trade set out at the accuracy ceiling unsupervised administration imposes. A study needing supervised administration cannot buy its way out with a cheaper platform. A study spanning languages carries a second constraint priced nowhere on this page, discussed at culture reduced instruments and what they trade away, and analysis planning should start from the percentile a composite converts to rather than from raw totals.

9 Scenario Three: A School District, Five Hundred Students

A district testing five hundred students with twenty five school psychologists spends $15,000 on the platform and $46,150 on staff time, for a modelled $122.30 per student. At this scale the volume bands finally work in the buyer's favor, and one of them produces an arithmetic quirk worth knowing.

Platform side: twenty five Standard Licenses in the twenty five or more band at $260 each, which is $6,500.00. Five hundred students at ten comprehensive subtests each is 5,000 subtests, which prepaid sits in the 751 to 5,000 band at $1.70, giving $8,500.00. Platform subtotal $15,000.00, or $30.00 per student.

Labor side: two hours per student at the school psychologist rate of $46.15 is $92.30 per student, or $46,150.00 across the district. Total $61,150.00, or $122.30 per student, of which 75.5 percent is time.

Now the quirk. The sheet lists $1.70 for 751 to 5,000 subtests and $1.45 for 5,001 or more. A district prepaying exactly 5,000 subtests pays $8,500.00. A district prepaying 5,001 pays $7,251.45. Buying one additional subtest appears to save $1,248.55, which is the sort of boundary that exists in any banded price list and that nobody notices because nobody does the arithmetic in public.

The district scenario also exposes the seat question. Twenty five seats at $260 is $6,500 whether those staff each test twenty students or forty. If half the assessment load sits with eight people, licensing twenty five seats buys availability rather than throughput, and counting who actually administers rather than who might is usually the largest single saving at this scale.

A district also carries obligations no cost model prices. The instrument must be defensible for the decision it informs, which means the evidence discussed at how measurement quality is actually evidenced before a placement, the threshold questions behind the score band used for gifted identification, and the fairness questions covered at whether the instrument functions the same way across groups. Getting those wrong costs far more than any license band.

10 The Platforms That Publish No Price at All

Cambridge Cognition and Creyos both decline to publish a rate card, so the only figures in circulation for them are directory listings neither vendor wrote. That distinction matters enough to state plainly rather than bury in a footnote.

Cambridge Cognition's departmental licences page for CANTAB Connect Research contains no price, no license duration, no user count and no tier structure. It contains an invitation to speak to an expert. The Capterra listing for CANTAB, read on September 3, 2026, shows a starting price of GBP 30,000 per user described as a one time charge, with a free trial available and no free version. That is a directory field supplied to a listing site, not a quote and not a published rate.

Creyos is the same shape at a very different magnitude. The Capterra listing for Creyos Health, read the same day, shows a starting price of $250.00 per month with a free trial and no free version. Creyos does not publish that on its own site in any form this page could verify, and a starting price tells you the floor of a range whose ceiling is unknown. The two directories carrying that floor do not even agree on what it includes, one marking a free version available and the other not, one of the checks run in the review of what a Creyos subscription actually buys, alongside the twelve task battery it covers and the four category scores it returns instead of a composite.

Treat both accordingly. A directory listing is a real, citable, dated data point about what a vendor told a listing site. It is not the price you will be quoted, it does not say what is included, and it cannot be reconciled against a competitor's rate card because the two are not the same kind of object. Comparing Pearson's $1.95 per subtest against GBP 30,000 per user is comparing a unit price against a contract.

The consequence for a buyer is procedural. When a vendor publishes nothing you cannot budget before the sales conversation, you cannot compare before disclosing your volume, and you have no public anchor when negotiating. Ask three questions in writing on the first call: what is the annual fee, what is the per administration fee, and what happens to both if our volume halves. A vendor who will not answer those in writing has told you something.

None of this makes a quoted platform worse. Research grade instruments with long validation histories are often sold as contracts because they are deployed as programs rather than bought as items, and language free batteries in particular, described at nonverbal batteries and what they can and cannot substitute for, are frequently sold that way. The argument here is narrower: the published and unpublished halves of this market cannot sit in one per unit table, which is why the directory pages ranking for this query contain no numbers, and why the questions at what accuracy claims a buyer should demand evidence for matter more than any quoted figure.

11 The Qualification Gate Is Also a Cost

You cannot buy the WAIS-5 without a doctorate, and that gate stops more organizations than any price on this page. Publisher qualification levels are usually discussed as an ethics topic. They are also, unavoidably, a procurement constraint.

Pearson's qualifications policy sets Level B as requiring a master's degree in psychology, education, speech language pathology, occupational therapy, social work or counseling, with formal training in ethical administration and interpretation, or equivalent licensure. Level C requires a doctorate in psychology, education or a closely related field with formal training in the ethical administration, scoring and interpretation of clinical assessments, or state licensure, or membership of bodies such as APA, NASP, NAN or INS. The WAIS-5 product page lists it as Qualification Level C.

Read that as a cost and the picture changes. A research team without a doctoral level psychologist cannot purchase the instrument at all, whatever the budget. Compliance means hiring or contracting that qualification, which at the BLS May 2025 median of $100,580 is an order of magnitude above the platform spend in every scenario above. Purchase eligibility is also only the first of three independent gates, because a state licensing board decides whether cognitive testing sits inside your scope of practice and the body receiving the report decides whether it accepts a score signed by someone with your credentials, a sequence set out at the guide to the qualification levels and who satisfies them. Clearing the publisher tells you nothing about the other two, and either can strand money already spent.

This is the real reason so many research programs run unsupervised or self administered batteries. It is rarely that the license fee was unaffordable; it is that the qualification, examiner availability and scheduling burden were, and the design fitting the constraint got chosen. Saying that out loud is better than presenting a logistics decision as a psychometric one.

Q-interactive's own tiering reflects the same logic from the other side. The classroom license for training partners is $150 a year for unlimited users and the experiential license is free, because those users are being trained rather than delivering services, and neither comes with the free starter kits a paying license includes.

If the gate blocks you, the remaining routes are worth mapping before you spend anything. Referral options are set out at where a qualified administration can actually be arranged. Adult assessment for identification purposes runs into the same wall, described at adult gifted assessment and who may administer it. And admission bodies maintain their own accepted instrument lists, catalogued at which instruments societies actually accept, which no amount of platform spending changes.

12 Where an Unsupervised Battery Changes the Arithmetic

Removing the examiner hour removes roughly three quarters of the cost, and roughly three quarters of the control at the same time. Both halves of that sentence have to be said together or the comparison becomes marketing.

ACIS Professional prices administrations in credits where one credit is one US dollar: 15 for a Quick administration, 30 for Optimized, 50 for Full Scale, and 5 per subtest for a custom battery. There is no seat fee, no kit, no consumable and no subscription requirement. Credits are charged when a participant link is created rather than when the participant starts, they do not expire, and a new account includes three Quick administrations at no cost so a team can pilot before committing budget. Monthly plans exist for recurring volume and stay optional.

Structurally that collapses five cost lines into one. Against the two hundred participant scenario, $50.00 per Full Scale administration against a modelled $116.82 is not a small difference, and the difference is almost entirely the examiner hour rather than any efficiency in the software. The same instrument sold to individuals is described at the consumer facing administration this workspace shares.

Now the boundary, stated as plainly as the price. ACIS is unsupervised and self administered. It is not a clinical instrument and not for diagnosis, hiring decisions, accommodation requests or high IQ society admission. Its reference frame is 3,243 English speaking records aged 16 to 90, self selected rather than census based, which is a genuine limit on how any score should be read. The technical analysis set for its confirmatory factor and reliability tables is 2,750 complete records.

Within those limits the measurement evidence is published rather than asserted. The Full Scale IQ composite omega is .9886 and its g loading is .958, with a standard error of measurement of about 1.60 IQ points, all on that self selected, unsupervised sample and a modelled adult reference frame rather than a census sample. Higher order model fit is CFI .9761, TLI .9726, RMSEA .0406, SRMR .0217. The structure is 20 subtests across six primary cognitive domains, arranged on the CHC hierarchy that defines those broad abilities, with scaled scores at mean 10 and standard deviation 3 and composites at mean 100 and standard deviation 15.

Every one of those numbers is derived in the ACIS technical manual, the reference frame method is set out at how a norm sample is built and what a self selected one costs you, and the interpretive limit on a loading of that size is explained at what a g loading does and does not license. Rarity claims derived from a composite should be checked against how rare a given score actually is before they enter a report.

13 Build Your Own Quote in an Afternoon

The method matters more than any number on this page, because every number here expires. Six steps produce a defensible internal cost per administration for whatever instrument and volume you actually have, and they take one afternoon.

Step one, fix the unit. Decide whether you are pricing a full battery, a single index or a screening administration, and write down exactly which subtests are in it. Almost every disagreement about cognitive testing cost is two people pricing different units, and the difference between a composite and its rank is a related confusion worth settling first at how a score differs from its percentile rank.

Step two, count seats honestly. Count the people who will personally administer at least one assessment in the license year, not those eligible to. This is the largest recoverable saving at institutional scale.

Step three, multiply the unit by volume and check the band. Banded prepay pricing rewards a slightly larger commitment and punishes an estimate landing just under a threshold. Get the band interpretation in writing.

Step four, price your own hour. Use your actual loaded staff cost rather than the bare wage used here, because a loaded rate is commonly a multiple of the wage and the ratio between labor and platform widens accordingly.

Step five, measure your minutes rather than assuming them. Time five real cases end to end, including scoring and report writing. The gap between assumed and measured time is routinely larger than the entire platform fee.

Step six, date the file and set a reminder to re-run it. Publisher list prices change at least annually; the Q-interactive figures here come from a sheet produced in November 2025 for the 2026 year, and a 2027 sheet will replace it.

Two constraints bound all of this. Cost narrows the options you can afford; it never justifies one. The Standards for Educational and Psychological Testing (2014), published jointly by the American Educational Research Association, the American Psychological Association and the National Council on Measurement in Education, require that the evidence supporting a test's use be appropriate to the specific decision being made, and that administration conditions match those under which the norms were collected. APA testing standards place that judgment on the professional selecting the instrument, not on the vendor who priced it. A cheaper route that changes the administration conditions has changed the interpretive claim you are entitled to make, and the predictive limits at what a cognitive score does and does not predict about performance bound that claim further. Price the options, then choose the one the standards allow, reading the result against exact percentile values for a finished composite rather than against the invoice.

14 Frequently Asked Questions