Tom Krause, NOSSCR Lifetime Member

From WAIS-5 Scores to Work-Related Limitations: What each Index Actually Predicts about Work and How to Say it in SSA’s own Vocabulary

Drawn from the author’s NOSSCR CLE training session scheduled for September 9, 2026 at 2pm. Register here.

Every representative in this practice has had the same experience. A neuropsychological report arrives, you read it, and it is obvious the claimant cannot work. Then you watch the ALJ shrug. The reason is almost never that the report was wrong. It is that the report stopped at the score and never continued to find a work limitation. This article is about continuing to the work-related limitations.

The bridge, and the three numbers that finish it

A score measures a construct under laboratory conditions: one problem at a time, in a quiet room, with a patient examiner. A limitation is a conclusion about function sustained across a workday and a workweek. Processing Speed Index of 74 is a test score. It implies slowed visual scanning, decision, and output; that is a construct. That implies difficulty holding pace on a quota task, a work demand. And the work demand must be phrased in the agency’s own vocabulary: the ability to complete a normal workday and workweek at a consistent pace.

But even a report that has established that an ability is impaired must describe how often, how severely, or for how long. Those three coordinates, frequency, intensity, and duration, are what convert a construct into a work-related limitation. SSA considers the kind, degree, and frequency of the difficulty a claimant would have (Listing 12.00F1); asks how often exacerbations occur and how long they last (Listings 12.00F4a); lists “location, duration, frequency, and intensity” of the pain among the required factors for symptom evaluation, 20 C.F.R. § 404.1529(c)(3); and describes the basic mental demands of unskilled work as the abilities, on a sustained basis, to do those things, POMS DI 25020.010A.3.a.

First, know what instrument you are looking at

The WAIS-5, published in 2024 with norms collected in 2023–2024, is not the WAIS-IV. The WAIS-5 has 20 subtests, 5 primary indexes where there were 4, and 15 ancillary indexes, but only 7 subtests feed the Full Scale IQ. A summary of the Indexes and subtests follows this article.

Perceptual Reasoning has split into Visual Spatial and Fluid Reasoning, so a claimant can now be intact on one and impaired on the other. Picture Completion and Cancellation are gone. Digit Span has been broken into three standalone subtests, which matters whenever compared to an older WAIS-IV report on the same claimant. Arithmetic now connects to (“loads on”) Fluid Reasoning rather than Working Memory.

The practical consequence is the one to carry into the hearing. Because the FSIQ leans toward higher-order reasoning and away from speed and working memory, a claimant can post a respectable Full Scale score and still be unable to keep pace. ALJs anchor on the FSIQ. Our job is to move them onto the index profile.

The five primary indexes, and the work each one predicts

What follows is the translation itself. For each index: what it measures, the work demands a low score bears on, and the POMS DI 25020.010B.3 abilities it puts in issue. Those abilities are the language the MRFC form and the written decision actually use, which is why the third step matters as much as the first two.

Verbal Comprehension (VCI). VCI measures acquired verbal knowledge and reasoning with words, built from Similarities and Vocabulary. A low VCI bears on understanding instructions, learning job terminology, written policies, and safety notices, and on asking a useful question when something goes wrong. It puts in issue the abilities to understand and carry out simple instructions, to ask simple questions or request assistance, and to be aware of normal hazards. Two cautions, both of which SSA may raise if you do not: First, VCI is heavily loaded on education and language, so a low score in a claimant with an eighth-grade education, or who learned English at age twenty-five, may be measuring opportunity rather than impairment. And second, an intact VCI does not rebut a brain-injury claim, because vocabulary is often the last thing to go.

Visual Spatial (VSI). VSI measuresanalyzing and mentally manipulating visual-spatial information, from Block Design and Visual Puzzles. A low VSI bears on blueprints, diagrams, and gauges; on assembly, packing, sorting, and machine set-up; and on wayfinding and hazard awareness. It puts in issue the abilities to remember work-like procedures, to carry out simple instructions, and to be aware of normal hazards and take appropriate precautions.

Fluid Reasoning (FRI). FRI involvessolving novel problems by inductive, quantitative, and logical reasoning, from Matrix Reasoning and Figure Weights. A low FRI bears on troubleshooting, adapting a routine, sequencing a multi-step job, and using judgment, and it puts in issue the abilities to make simple work-related decisions and to respond appropriately to changes in a routine work setting. Note what FRI does not establish. It is measured one problem at a time, in a quiet room, with a patient examiner, so it speaks to reasoning under ideal conditions rather than reasoning under production pressure with a phone ringing.

Working Memory (WMI). WMI gaugesholding and manipulating information over seconds while using it, from Digit Sequencing and Running Digits. A low WMI bears on multi-step instructions, resuming after an interruption, taking an order, and mental arithmetic. It puts in issue the abilities to carry out simple instructions and to maintain attention for extended periods of two-hour segments. That second ability is where a great many of these cases are actually decided.

Processing Speed (PSI). PSI calculatesthe speed of simple visual scanning, decision, and output under time limits, from Coding and Symbol Search. A low PSI bears on production quotas, timely completion, keeping pace with coworkers on routine output, and error rate late in a shift, and it puts in issue the ability to complete a normal workday and workweek at a consistent pace. PSI is the most sensitive index in the battery to various causes and the least specific. Depression, pain, sedating medication, sleep loss, poor effort, and even a bad shoulder all depress it. A low PSI tells you that the claimant is slow; it never tells you why, but a good report answers that question before the ALJ asks it.

The ancillary indexes

These are the most underused tools in the instrument, and two of them help decide cases. General Ability (GAI) strips out speed and working memory and gives you the claimant’s ceiling. Cognitive Proficiency (CPI) strips out reasoning and gives you her efficiency. The gap between them may be the basis for the entire claim, because the strongest psychometric case is rarely that the client is not smart. It is that the client is smart and still cannot keep up.

The motor-reduced family matters in any case with a severe physical impairment. Block Design is timed and requires two working hands, so in a claimant with tremor, neuropathy, arthritis, or hemiparesis a low score may be measuring the hand rather than the brain. The Nonmotor Index and the Motor-Reduced Processing Speed Index estimate ability without the manual demand, and the new Naming Speed Quantity subtest measures motor-free numerical fluency. Together they are the answer to “that is just her arthritis.” If your client has an upper-extremity problem, the first question to the neuropsychologist is whether a motor-reduced composite was reported; if not, request it. The Nonverbal Index does the same work where aphasia, expressive language loss, or limited English proficiency would otherwise confound the estimate.

The auditory working memory split is also more useful than it looks. Registration measures taking information in; Manipulation measures working with it while holding it. A claimant who registers normally but does poorly on manipulation can hear an instruction and still be unable to act on a compound one. That is a specific and credible limitation, and a far better sentence than “impaired working memory.”

Subtests are hypotheses, not conclusions

Argue indexes, not subtests. An index is a composite and far more reliable than any single scaled score, and across a twenty-subtest battery healthy adults routinely produce at least one low score. A subtest-level finding tells you what a score is consistent with, not what it proves, and it never tells you how often, how severely, or for how long. A low PSI alone is a number. Convergence is the whole game: a low PSI plus a slowed Trails A plus an employer who documents missed quotas is an argument.

Two subtest-level moves are worth knowing anyway. First, structure dependence: an intact Block Design, where a model is supplied, beside a low Matrix Reasoning, where none is, says the claimant performs better when the task has been structured for her. This speaks directly to responding appropriately to changes in a routine work setting. Second, Spatial Addition is the only WAIS-5 subtest that expressly requires ignoring competing stimuli, which makes it the closest the instrument comes to the conditions of an actual workplace and the best evidence in the battery on the ability to work in coordination with or proximity to others without being unduly distracted by them.

Governing all of it is the instructions found at Listing 12.00F3d: no single piece of information, including test results, can establish the degree of limitation of an area of mental functioning. This cuts both ways. It is also your answer when a denial rests on one 15-minute consultative examination that found the claimant alert and oriented times three.

From an index to a hypothetical question

The test of a translated finding is whether it could be read into a hypothetical. “Processing speed is moderately impaired” cannot be: it is true, it is defensible, and it is useless, because an ALJ can accept it in full and still write an RFC for simple, routine tasks. This can be: a Processing Speed Index of 74 places the claimant at roughly the fourth percentile for simple visual output; convergent with her working memory scores (Ex. __F/__), a slowed Trails B (Ex. __F/__), her decline across the continuous performance measure (Ex. __F/__), and her employer’s quota records (Ex. __E/__), she would be off task for 15-20% of an eight-hour day. (Ex. __F/__).

Note that the second version names its sources, and that this is not a stylistic preference. Usually, a neuropsychological evaluation is read by a judge outside of the hearing. The ALJ must then explain in writing how persuasive she found it. Under 20 C.F.R. § 404.1520c(b)(2) the two factors she must address are supportability and consistency, and supportability is defined at (c)(1) by the objective evidence and the supporting explanations the source presents. So ask the expert for the reasoning rather than only the number, and for a range rather than false precision. For example, 15-20% survives a skeptical reading in a way that 17% does not.

Then take the number to the vocational expert. Ask what percentage of the workday off task is tolerated in the unskilled jobs she identified and ask for the basis for the figure. Ask how many unscheduled absences a month are tolerated. Ask it twice: once for the general tolerance and once for the probationary period. See Sczepanski v. Saul, 946 F.3d 152 (2d Cir. 2020) (claimant’s ability to complete a probationary period was relevant to her disability status). These tolerances vary by expert and by case and are testimony rather than authority, so establish yours on the record before you argue that your client exceeds it.

What the WAIS-5 cannot reach

At least one gap remains. Working memory on the WAIS-5 is a seconds-long construct; that is, a digit sequence held and reordered. Attention on a continuous performance test is minutes. The POMS abilities live an order of magnitude further out: attention across two-hour segments, a full shift at a consistent pace. Regular attendance and punctuality, eight hours a day and five days a week, and the twelve-month duration requirement live further out still. Every instrument in the report sits in the first two registers; every ability that decides the case sits in the last two. That is also why the WAIS-5 alone often is not enough — memory, executive function, sustained attention, adaptive functioning, and validity testing are all found elsewhere.

No test closes the gap; the record does. Obtain employer and personnel records, including attendance data, write-ups, quota reports, accommodations granted, and the reason the job ended. These records may be the best evidence of the inability to sustain a workweek anyone will put before the ALJ. And often, the most overlooked source is the neuropsychological evaluation itself. The neuropsychologist watched your client for perhaps 2-3 hours and knows how many breaks she took, how often an instruction had to be repeated, and whether the last hour looked like the first. That is frequency and duration data observed by an acceptable medical source, and it is often omitted because nobody asked for it in writing.

An example

A 47-year-old shipping clerk, three years past a moderate TBI. VCI 104, VSI 96, FRI 99, WMI 78, PSI 74, FSIQ 88, with GAI 101 and CPI 76. Premorbid FSIQ estimated near 102, a flat CVLT-3 learning curve, slowed Trails B, all validity measures passed. Read the indexes and the translation writes itself: reasoning is preserved, efficiency has collapsed, and the claim runs to maintaining attention across two-hour segments and completing a normal workday and workweek at a consistent pace. The claimant’s limitations are quantified, on this record, at roughly 15% off task, worse late in each segment, and 1 ½ to 2 unscheduled absences a month. What it does not support is substantial loss of the ability to understand or carry out simple instructions, where VCI and FRI are intact. Reaching for that lets the ALJ discount everything else. Claim attention and pace, not substantial loss.

If you remember nothing else

Move the ALJ off the FSIQ and onto the index profile. Argue indexes rather than subtests, and convergence rather than single findings. Learn the GAI–CPI split, because it is the profile that wins. Correlate every index to the POMS DI 25020.010B.3 abilities by name, one at a time. And give every limitation a frequency, an intensity, and a duration with a stated basis and cite, because an ability answered yes or no is an ability the ALJ cannot put to a vocational expert.

WAIS-5 Subtests

The Wechsler Adult Intelligence Scale-Fifth Edition (WAIS-5) is divided into 10 Primary subtests (used for core IQ metrics) and 10 Secondary subtests (used for deeper clinical insight) across five main domains.

Verbal Comprehension Subtests

• Vocabulary (Primary): The examinee defines words presented orally and visually.
• Similarities (Primary): The examinee explains how two seemingly different words or concepts are alike.
• Information (Secondary): The examinee answers general knowledge questions covering history, science, and geography.
• Comprehension (Secondary): The examinee answers questions about social rules, common expressions, and everyday solutions.

Visual Spatial Subtests

• Block Design (Primary): The examinee uses red-and-white plastic blocks to replicate a printed geometric design within a time limit.
• Visual Puzzles (Primary): The examinee chooses three puzzle pieces from an array that can combine to recreate a target shape.

Fluid Reasoning Subtests

• Matrix Reasoning (Primary): The examinee completes an incomplete visual grid by selecting the logical missing piece from multiple choices.
• Figure Weights (Primary): The examinee selects weights to balance an imperfect scale based on visual logic diagrams.
• Set Relations (Secondary): The examinee identifies patterns and deductive relationships among abstract visual shapes.
• Arithmetic (Secondary): The examinee mentally solves mathematical word problems without using a pencil or calculator.

Working Memory Subtests

• Digit Sequencing (Primary): The examinee listens to a jumbled string of numbers and repeats them back in exact numerical order.
• Running Digits (Primary): The examinee listens to a continuous string of numbers and must remember and repeat only the last few digits spoken.
• Digits Forward (Secondary): The examinee listens to a sequence of numbers and repeats them back exactly as heard.
• Digits Backward (Secondary): The examinee listens to a sequence of numbers and repeats them in reverse order.
• Symbol Span (Secondary): The examinee views a brief sequence of abstract symbols and then selects them in the correct sequential order from a grid.
• Spatial Addition (Secondary): The examinee views two sequential visual grids and mentally combines or subtracts the location of dots to select a final pattern.
• Letter-Number Sequencing (Secondary): The examinee listens to mixed letters and numbers and reorders them alphabetically and numerically. 

Processing Speed Subtests

• Coding (Primary): The examinee copies simple symbols that pair with specific numbers as quickly and accurately as possible within a time limit.
• Symbol Search (Primary): The examinee scans a visual row to quickly determine if a target symbol matches an array of options.
• Naming Speed Quantity (Secondary): The examinee rapidly looks at visual groups of boxes and vocally names the quantities of those boxes as fast as possible.

---

The views and opinions expressed in this article are the author’s own and do not necessarily reflect the views of NOSSCR, its leadership, or staff.

---