Skip to main content
v2026.11,986 entries · CC-BY 4.0

Symbol Digit Modalities Test: SDMT Scoring Guide

The Symbol Digit Modalities Test (SDMT) measures cognitive processing speed. Learn administration protocols, z-score formulas, and clinical cutoffs.

Written and maintained by CASRAI Editorial Board

Published · Last updated

On this page

Methodological Note: This technical reference documents standardized administration protocols, psychometric equations, and normative interpretation criteria for cognitive research. It provides methodological documentation and does not constitute individual clinical advice.

Definition in One Paragraph

The Symbol Digit Modalities Test (SDMT) is a standardized neuropsychological instrument developed by Aaron Smith that measures cognitive processing speed, visual scanning efficiency, sustained attention, and visual-motor coordination. Examinees refer to a displayed key pairing nine abstract geometric symbols with the digits 1 through 9, rapidly substituting the corresponding number for randomized symbol sequences within a strict 90-second testing window. The test is administered in either written form, where respondents write numbers beneath symbols, or oral form, where respondents speak the digits aloud while an examiner records them. Because the substitution task demands minimal linguistic processing and relies on universal geometric symbols and basic digits, the SDMT provides a culture-reduced assessment tool widely used in cognitive neuroscience and clinical trials, serving as the benchmark screening instrument for cognitive decline in multiple sclerosis (MS) and closed-head injury. For a broader bedside cognitive screen, see the Montreal Cognitive Assessment (MoCA).

Formula with Every Symbol Explained

Evaluating SDMT performance requires converting raw substitution totals into standardized normative scores and determining whether longitudinal changes exceed psychometric error.

Raw Score and Standardized Conversions

The primary index is the raw score (X), representing total correct substitutions completed in 90 seconds (excluding 10 practice items). Raw scores are converted to a standardized z-score to compare performance against a demographically matched reference sample:

z = (X – μ) / σ

Where:

  • X is the observed raw score (count of correct substitutions in 90 seconds).
  • μ is the normative sample mean matched for age group, educational level, and administration format (oral or written).
  • σ is the standard deviation of the normative reference sample.
  • z represents the standard score in standard deviation units relative to the population mean.

The z-score is converted to a conventional clinical T-score via:

T = 50 + (10 × z)

Standard Error of Measurement and Reliable Change Index

In clinical trials, investigators determine whether longitudinal score shifts reflect true disease progression or random test-retest error. The Standard Error of Measurement (SEm) is:

SEm = σ × √(1 – rxx)

Where rxx is the test-retest reliability coefficient (typically rxx = 0.80 to 0.86). The standard error of the difference across two testing waves is:

SEdiff = √(2 × SEm2)

To verify if individual decline is statistically reliable while adjusting for control practice effects (μpractice), researchers calculate the Reliable Change Index (RCI):

RCI = [(X2 – X1) – μpractice] / SEdiff

Where X1 is baseline score, X2 is follow-up score, and |RCI| ≥ 1.645 confirms reliable change at the 90% confidence level (two-tailed p < 0.10).

Worked Example with Real Numbers and the Arithmetic Shown

Consider a longitudinal trial monitoring cognitive processing in a 42-year-old participant with 14 years of formal education evaluated with the oral SDMT.

Step 1: Baseline Assessment and Normalization

At baseline (Time 1), the participant achieves a raw score of X1 = 52. Published normative parameters for adults aged 40–44 with 13–15 years of education are μ = 58.4, σ = 7.2, and rxx = 0.84.

z1 = (52 – 58.4) / 7.2 = -6.4 / 7.2 = -0.8889 ≈ -0.89

T1 = 50 + [10 × (-0.8889)] = 50 – 8.889 = 41.111 ≈ 41.1

Because z1 = -0.89 falls within one standard deviation of the normative mean, baseline processing speed is intact.

Step 2: 12-Month Follow-Up and Raw Difference

At Month 12 (Time 2), the participant achieves a follow-up raw score of X2 = 46.

ΔX = X2 – X1 = 46 – 52 = -6 points

% Change = [(46 – 52) / 52] × 100% = (-6 / 52) × 100% = -11.54%

In multiple sclerosis monitoring, a decline of ≥4 raw points (or ≥10%) satisfies the consensus benchmark for clinically meaningful cognitive decline (Benedict et al., 2017).

Step 3: Reliable Change Index (RCI)

Assuming an expected 12-month control practice gain of μpractice = +1.5 points:

SEm = 7.2 × √(1 – 0.84) = 7.2 × √(0.16) = 7.2 × 0.40 = 2.88

SEdiff = √(2 × 2.882) = √(2 × 8.2944) = √16.5888 ≈ 4.0729

RCI = [(-6) – 1.5] / 4.0729 = -7.5 / 4.0729 ≈ -1.84

Because |-1.84| exceeds the 90% confidence threshold (|z| ≥ 1.645), this 6-point decline represents statistically reliable cognitive deterioration rather than measurement error.

Interpretation: What the Result Means

Evaluating SDMT performance requires synthesizing static cross-sectional scores and longitudinal trajectories, distinguishing between statistical significance and clinical significance.

Range Z-Score T-Score Clinical Interpretation
Above Average z ≥ +1.00 T ≥ 60 Superior information processing speed and visual scanning efficiency.
Intact / Normal -1.00 < z < +1.00 40 < T < 60 Normal performance within one standard deviation of demographic peers.
Mild Slowing -1.50 < z ≤ -1.00 35 < T ≤ 40 Subtle reduction in processing efficiency; warrants serial observation.
Impaired z ≤ -1.50 T ≤ 35 Standard screening cut-point for cognitive deficit in MS clinical research.
Longitudinal Progression ΔX ≤ -4 raw pts ≥10% decline Consensus threshold for clinically meaningful cognitive decline (Benedict et al., 2017).

Modality Discrepancy: Oral Versus Written Administration

Dual administration formats provide built-in differential diagnostic capability. When upper-extremity motor impairment or tremor is present, written testing underrepresents cognitive speed. An oral score exceeding written performance by more than 1.5 standard deviations (zoral – zwritten > 1.5) confirms that motor execution barriers, rather than central cognitive slowing, explain the deficit.

Clinical Application: SDMT in Multiple Sclerosis (SDMT MS)

Information processing speed is the cognitive domain most vulnerable to cerebral white matter demyelination in multiple sclerosis. International consensus batteries (BICAMS and MACFIMS) mandate the oral SDMT as the primary cognitive outcome measure. Clinical trials utilize a confirmed 4-point raw decline sustained over 3 to 6 months as an objective secondary endpoint for disease progression.

Assumptions and Common Mistakes

Valid interpretation requires adherence to standardized test administration:

  • Assumption of Adequate Vision: The test requires distinguishing subtle geometric shapes. Testing individuals without corrective eyewear or during active optic neuritis produces invalid normative comparisons.
  • Practice Effects in Short-Interval Retesting: Examinees develop associative familiarity across repeated trials. Testing intervals under 90 days without practice-adjusted RCIs or alternate forms inflate scores by 3 to 6 points.
  • Timing Non-Compliance: The 90-second limit must be enforced with a stopwatch. Adding even five seconds inflates scores by 2 to 4 substitutions, obscuring mild deficits.
  • Distinction from the Trail Making Test: While both measure processing speed, the Trail Making Test (TMT) introduces spatial sequencing (Part A) and executive task-switching (Part B), whereas the SDMT isolates sustained pairing and visual scanning. They serve complementary functions and are not interchangeable.

How to Report It (APA Example)

Reports following APA 7th style must document administration modality, normative dataset, baseline and follow-up metrics, standardized scores, and reliable change determinations.

Example APA 7th Narrative:

“Information processing speed was evaluated using the oral Symbol Digit Modalities Test (SDMT; Smith, 1982). At baseline, the participant completed 52 correct substitutions within the standardized 90-second trial, corresponding to a z-score of -0.89 (T = 41.1) relative to age- and education-stratified norms. At 12-month follow-up, the participant completed 46 substitutions (ΔX = -6 points, -11.5% change), satisfying the consensus threshold for clinically meaningful cognitive decline (≥4 points; Benedict et al., 2017). Calculation of the Reliable Change Index adjusting for 12-month practice effects (μpractice = +1.5, rxx = 0.84, SEdiff = 4.07) confirmed statistically reliable decline, RCI = -1.84, p < .10.”

How to Calculate It in Excel, R, and Python

The following verified code snippets implement SDMT scoring, z-score standardization, and Reliable Change Index calculations.

Microsoft Excel Formulas

Assuming baseline data in row 2 with Raw Score (X) in B2, Normative Mean (μ) in C2, Normative Standard Deviation (σ) in D2, Reliability (rxx) in E2, Follow-up Score in F2, and Practice Gain in G2:

  • Z-Score: =(B2 - C2) / D2
  • T-Score: =50 + (10 * ((B2 - C2) / D2))
  • Standard Error of Difference: =SQRT(2 * ((D2 * SQRT(1 - E2))^2))
  • Adjusted RCI: =((F2 - B2) - G2) / SQRT(2 * ((D2 * SQRT(1 - E2))^2))

R Implementation

calc_sdmt <- function(t1, t2 = NULL, m = 58.4, s = 7.2, r = 0.84, p = 1.5) {
  z1 <- (t1 - m) / s
  se_diff <- sqrt(2 * (s * sqrt(1 - r))^2)
  res <- list(raw_t1 = t1, z1 = round(z1, 2), t1_score = round(50 + 10 * z1, 1))
  if (!is.null(t2)) {
    d <- t2 - t1
    res$raw_t2 <- t2
    res$delta  <- d
    res$rci    <- round((d - p) / se_diff, 2)
    res$decline <- d <= -4
  }
  return(as.data.frame(res))
}
calc_sdmt(52, 46)

Python Implementation

import math

def calculate_sdmt(t1: int, t2: int = None, m: float = 58.4, s: float = 7.2, r: float = 0.84, p: float = 1.5) -> dict:
    z1 = (t1 - m) / s
    se_diff = math.sqrt(2.0 * ((s * math.sqrt(1.0 - r)) ** 2))
    res = {"raw_t1": t1, "z1": round(z1, 2), "t1": round(50.0 + 10.0 * z1, 1)}
    if t2 is not None:
        d = t2 - t1
        res.update({"raw_t2": t2, "delta": d, "rci": round((d - p) / se_diff, 2), "decline": d <= -4})
    return res

if __name__ == "__main__":
    print(calculate_sdmt(52, 46))

Frequently Asked Questions

What is the Symbol Digit Modalities Test (SDMT) primarily used for in Multiple Sclerosis (SDMT MS)?

In multiple sclerosis research, the oral SDMT serves as the benchmark screening instrument for cognitive processing speed, the cognitive domain most sensitive to white matter demyelination and cerebral atrophy. Recommended by international consensus panels including BICAMS and MACFIMS, the SDMT detects processing speed decline earlier and with higher sensitivity than broader test batteries.

How does the Symbol Digit Modalities Test compare to the Trail Making Test?

While both instruments measure cognitive efficiency, the SDMT assesses rapid visual scanning and paired-associate substitution from a fixed key, whereas the Trail Making Test evaluates visual-motor search (Part A) and executive cognitive flexibility and set-shifting (Part B). They assess distinct cognitive components and are frequently paired together in comprehensive neuropsychological batteries.

Can researchers access a Symbol Digit Modalities Test PDF or free download?

The Symbol Digit Modalities Test is a copyrighted psychological assessment published exclusively by Western Psychological Services (WPS). While academic papers summarize administration guidelines and scoring equations, standardized test protocols, record sheets, and normative tables cannot be lawfully obtained as a free PDF download. Clinical and research investigators must purchase official AutoScore test forms directly from WPS or authorized assessment distributors.

What raw score change indicates true cognitive decline versus practice effect?

In multiple sclerosis trials, a reduction of 4 or more raw points (or ≥10% relative decline) is the validated consensus threshold for clinically meaningful cognitive deterioration (Benedict et al., 2017). To evaluate individual trajectory significance, researchers calculate the Reliable Change Index (RCI), which adjusts for test-retest error variance and expected retest practice effects.

Sources

  • Benedict, R. H. B., DeLuca, J., Phillips, G., LaRocca, N., Hudson, L. D., Rudick, R., & Multiple Sclerosis Outcome Assessments Consortium. (2017). Validity of the Symbol Digit Modalities Test as a cognition performance outcome measure for multiple sclerosis. Multiple Sclerosis Journal, 23(5), 721–733. https://doi.org/10.1177/1352458517690821 (PMID: 28206827; read 2026-11-09).
  • Langdon, D. W., Amato, M. P., Brombacher, J., de Luca, J., Geurts, J. J., Heesen, C., Penner, I. K., Rovira, A., Sanfilippo, C., Sastre-Garriga, J., & Benedict, R. H. (2012). Recommendations for a Brief International Cognitive Assessment for Multiple Sclerosis (BICAMS). Multiple Sclerosis Journal, 18(6), 891–898. https://doi.org/10.1177/1352458511431076 (PMID: 22190573; read 2026-11-09).
  • Parmenter, B. A., Weinstock-Guttman, B., Garg, N., Munschauer, F., & Benedict, R. H. (2007). Screening for cognitive impairment in multiple sclerosis using the Symbol digit Modalities Test. Multiple Sclerosis Journal, 13(1), 52–57. https://doi.org/10.1177/1352458506070750 (PMID: 17294611; read 2026-11-09).
  • Smith, A. (1982). Symbol Digit Modalities Test (SDMT): Manual (Revised). Western Psychological Services. Official publisher documentation: https://www.wpspublish.com/sdmt-symbol-digit-modalities-test (read 2026-11-09).
  • SpLD Assessment Standards Committee (SASC). (2022). Updated Guidance on the Use of the Symbol Digit Modalities Test. SASC Test Evaluation Committee Notice (read 2026-11-09).

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Symbol Digit Modalities Test: SDMT Scoring Guide

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →