Trail Making Test
The Trail Making Test (TMT) is a brief, standardized neuropsychological assessment instrument used to evaluate an individual's visual attention, processing speed, executive functioning, and cognitive flexibility. It consists of two primary parts administered on paper: Part A requires the participant to connect 25 consecutively numbered circles (1 to 25) as quickly as possible without lifting the pencil, while Part B demands connecting circles in an alternating sequence of numbers and letters (1-A-2-B-3-C, up to 13 numbers and letters). Completion times for each part are recorded, with errors gently corrected during administration, and the test is discontinued after five minutes if unfinished; normative data indicate average times of approximately 29 seconds for Part A and 75 seconds for Part B in healthy adults, though these vary by age, education, and population.[1][2][3] Originally developed in 1938 by psychologists John E. Partington and Russell G. Leiter as a measure of divided attention—initially called the Divided Attention Test or Partington's Pathways—the TMT was incorporated into the U.S. Army Individual Test Battery in 1944 to screen recruits for cognitive fitness during World War II. It gained prominence in clinical neuropsychology through refinements by Ralph M. Reitan in the 1950s, who integrated it into the Halstead-Reitan Neuropsychological Battery as a sensitive indicator of brain dysfunction, with Part B particularly highlighting frontal lobe impairments. Since then, the TMT has been translated into numerous languages, adapted for digital formats, and extensively normed across diverse populations, including children, older adults, and clinical groups, solidifying its role as a public-domain tool in both research and practice.[2][3][4] In interpretation, performance on the TMT is analyzed through raw completion times, error rates, and derived metrics such as the B/A ratio or difference score (TMT-B minus TMT-A), which help isolate executive deficits from basic visuomotor slowing; elevated times on Part A suggest attentional or perceptual issues, while disproportionate delays on Part B indicate problems with set-shifting and working memory. The test's sensitivity to conditions like traumatic brain injury, stroke, Alzheimer's disease, Parkinson's disease, and attention-deficit/hyperactivity disorder (ADHD) makes it valuable for differential diagnosis, treatment planning, and monitoring cognitive decline, though it must be interpreted alongside other assessments due to influences from motor speed, education, and cultural factors.[1][3][5] Beyond traditional clinical settings, the TMT informs broader applications, such as evaluating fitness to drive in older adults or assessing mild cognitive impairment in primary care; its brevity (typically under 10 minutes), low cost, and robust psychometric properties— including high test-retest reliability (r > 0.80) and validity correlations with other executive measures—contribute to its enduring popularity in neuropsychological batteries worldwide. Ongoing research explores computerized versions to enhance precision in tracking sub-movements and eye-tracking integrations for deeper insights into cognitive processes.[6][7][4]Overview
Description
The Trail Making Test (TMT) is a neuropsychological assessment that requires participants to connect a series of dots in a specific sequence on a worksheet, evaluating visual search, scanning, speed of processing, mental flexibility, and executive functioning.[1][8] This test serves as a measure of cognitive efficiency through tasks demanding sustained attention and task-switching between stimuli.[9] The TMT consists of two primary parts presented on separate sheets. In Part A, examinees draw lines to connect 25 circles labeled with numbers from 1 to 25 in ascending order. In Part B, they connect 25 circles alternating between numbers (1 to 13) and letters (A to L), following the sequence 1-A-2-B-3-C, and so forth.[1][8] The circles are scattered randomly across an 8.5 by 11-inch page to simulate real-world visual search demands without providing directional cues.[8][10] Designed as a quick, non-verbal tool, the TMT typically requires 5 to 10 minutes for administration, making it suitable for clinical settings where efficient screening of cognitive impairments is needed.[8][4]Purpose and Cognitive Domains
The Trail Making Test (TMT) serves primarily as a screening tool for detecting cognitive impairment in clinical and research settings, particularly in populations at risk for neurological conditions. It evaluates key aspects of brain function, including attention, processing speed, visual-motor coordination, and executive functions such as set-shifting. Developed as a brief, sensitive measure, the TMT is widely employed to identify deficits that may indicate underlying brain pathology, with its simplicity allowing integration into broader neuropsychological batteries.[11][12] In terms of cognitive domains, the TMT targets visual attention through the requirement to scan and locate stimuli amid distractors, psychomotor speed via the timed motor response of connecting targets, and cognitive flexibility in alternating between conceptual sets. Part A emphasizes basic visuomotor tracking and sustained attention, while Part B incorporates divided attention and working memory to maintain the alternating sequence (e.g., numbers to letters). These elements collectively probe fluid intelligence and inhibitory control, with performance influenced by age and education-related factors.[13][14][15] Clinically, the TMT's rationale lies in its differential demands: completion time for Part A reflects foundational processing speed and visual scanning efficiency, whereas Part B introduces executive demands that reveal set-shifting impairments when the B-A difference is elevated. This contrast helps isolate pure executive deficits from general slowing, aiding diagnosis in conditions like mild cognitive impairment where early executive changes are prominent. The test is particularly sensitive to frontal-subcortical circuit disruptions, as seen in traumatic brain injury and dementia, where prolonged times or errors signal impaired planning and flexibility.[15][14][16]History and Development
Origins
The Trail Making Test originated from the 1938 Divided Attention Test (also known as Partington's Pathways Test) developed by psychologists John E. Partington and Russell G. Leiter as a measure of divided attention through connecting pathways.[2][17] This test drew from earlier visual-motor and attention tasks and was incorporated into the U.S. Army Individual Test Battery (AITB) in 1944 during World War II to assess general mental ability and screen for cognitive impairments in military personnel, particularly recruits, in a non-verbal manner to ensure accessibility across diverse individuals.[18][19] First published in 1944 within the AITB manual, the test underwent early validation through its deployment in the U.S. Army for identifying neurological impairments, providing initial evidence of its utility in detecting brain damage via timed performance on sequential connections.[19] This wartime application established the TMT as a practical tool for rapid cognitive screening in high-stakes clinical settings.Standardization and Evolution
In the 1950s, Ralph M. Reitan refined and standardized the Trail Making Test (TMT) for clinical use within the Halstead-Reitan Neuropsychological Battery (HRNB), establishing normative data for individuals aged 15 to 79 to assess brain dysfunction in adults.[18] This integration emphasized the TMT's sensitivity to organic brain damage, with Reitan's manual providing guidelines for administration, scoring, and interpretation based on empirical validation against neurological criteria. Key evolutions in the 1990s involved updates by Otfried Spreen and Esther Strauss, who compiled and refined normative data in their seminal compendium, first published in 1991 (with subsequent editions in 1998 and 2006), extending applicability to broader demographic groups by incorporating adjustments for age, education, gender, and cultural factors to enhance interpretative accuracy.[20] Their work synthesized existing studies into a comprehensive framework, promoting the TMT's adoption in diverse neuropsychological assessments beyond the original HRNB context. The test was also incorporated into automated batteries like the Automated Neuropsychological Assessment Metrics (ANAM), where it or analogous tasks supported computerized evaluation of attention and executive functions in military and clinical settings. Normative expansions in the 2000s addressed gaps in pediatric and elderly populations, with the Comprehensive Trail Making Test (CTMT) by Cecil R. Reynolds providing stratified data for ages 9 to 89, enabling age-appropriate assessments in developmental and geriatric contexts. This development improved the TMT's utility for detecting cognitive impairments across the lifespan, filling post-1950s limitations in the original adult-focused norms. By the 2010s and into 2025, the TMT evolved toward digital integration, with platforms like the Vienna Test System's TMT-L offering automated administration, precise timing, and enhanced error analysis for improved reliability in research and clinics. Cross-cultural adaptations proliferated, including non-English versions such as the Color Trails Test, which substitutes colors for numerals to reduce literacy biases in diverse populations.Administration
Materials and Procedure
The Trail Making Test (TMT) requires minimal materials for administration, consisting of standardized worksheets printed on 8.5 x 11-inch paper, a pencil or pen, a stopwatch, and practice sheets for demonstration. Each worksheet features 25 circles: Part A contains numbered circles from 1 to 25 randomly distributed across the page, while Part B includes numbered circles from 1 to 13 interspersed with lettered circles from A to L. As a public-domain instrument, forms can be reproduced, but standardized versions are available from publishers like Psychological Assessment Resources, Inc., ensuring consistent layout and sizing for reliable testing.[21][22] Administration occurs individually in a quiet, distraction-free environment, such as a testing room with a flat table surface, to minimize external influences on performance. The examiner positions the practice sheet flat in front of the participant, provides a pencil, and delivers clear verbal instructions: for the general task, the participant is directed to connect the circles in sequential order as quickly as possible without lifting the pencil from the paper or going outside the circles. A demonstration using the sample sheet is provided, during which the examiner traces the path while verbalizing the sequence (e.g., "Now, number 1... number 2...") to model the process; the participant then practices under supervision, with the examiner offering immediate feedback if needed. Once the practice is complete, the formal test sheet is placed in the same orientation, and timing begins precisely when the participant starts drawing upon the command "Go," continuing until completion or a 5-minute (300-second) cutoff per part.[22][23] If the participant makes an error, such as connecting to the wrong circle or perseverating on a previous sequence, the examiner immediately intervenes by saying "Stop" to halt progress, redirects to the last correct connection, and resumes timing without pausing the stopwatch; this ensures the total time reflects the full effort, including corrections, while preventing uncorrected errors from invalidating the trial. After initiating the task, the participant is instructed not to speak or provide verbal commentary to maintain focus. The entire TMT, including both parts and demonstrations, typically takes 5 to 10 minutes to administer. Specific sequencing instructions differ between Parts A and B, as detailed in their respective sections.[22][23]Parts A and B
The Trail Making Test is divided into two distinct parts, A and B, each presented on a separate worksheet containing 25 circles distributed across an 8.5 by 11-inch sheet of paper.[8] Part A requires the participant to draw lines connecting circles numbered 1 through 25 in ascending sequential order as quickly and accurately as possible, using a pencil without lifting it from the paper. The numbers are positioned randomly to necessitate visual scanning across the page. This part emphasizes straightforward sequential linking of numerical targets.[1] Part B builds on this by having the participant connect 25 circles that alternate between numbers (1 through 13) and letters (A through L), in the order 1-A-2-B-3-C and so on up to 13-L, again as quickly and accurately as possible without lifting the pencil. The targets are similarly randomized, requiring shifts between identifying and linking numerical and alphabetic elements.[8] The primary difference lies in task structure: Part A follows a single, linear numerical progression, whereas Part B incorporates alternating sequences that demand set-shifting between two categories, leading to higher error rates and longer durations. Per standard protocols, Part B is typically discontinued after 5 minutes if the participant has not finished.[1][8]Scoring and Interpretation
Calculation of Scores
The primary scores for the Trail Making Test (TMT) are the completion times for Parts A and B, measured in seconds from the examiner's "go" signal until the participant draws a line connecting the final target (number 25 for Part A or the last number-letter pair for Part B).[8] These times reflect visuomotor speed and attention for Part A and additional set-shifting demands for Part B, with longer durations indicating poorer performance.[24] Part B is discontinued after 5 minutes (300 seconds) if not completed, with the elapsed time recorded as the score; Part A is typically completed within 90 seconds, but longer times are allowed and recorded as is.[8][23] Error scoring supplements the time-based primary metric but is secondary in quantitative analysis. Errors include connecting to an incorrect target (sequence breaks), repeating a previously connected target (perseverations), or other deviations such as off-trail lines, and are tallied separately for each part while the participant corrects them under examiner guidance without pausing the timer.[14] The time spent correcting errors is incorporated into the overall completion time, emphasizing efficiency over error count alone.[23] Qualitative observations, such as whether the participant lifts the pencil between connections (indicating scanning strategy) versus drawing continuous lines, may also be noted to inform behavioral insights, though these do not contribute to formal scores.[14] Derived metrics adjust raw times to better isolate cognitive components beyond basic motor speed. The B-A difference score subtracts the Part A completion time from the Part B completion time, providing a measure of executive dysfunction relatively independent of visual scanning and motor confounds: \text{B-A} = \text{TMT-B time} - \text{TMT-A time} [8] This subtraction helps attribute performance decrements in Part B to cognitive flexibility rather than general processing speed.[24] Additionally, the B/A ratio divides the Part B time by the Part A time, offering a relative index of set-shifting efficiency that normalizes for individual differences in baseline speed.[8]Normative Data and Cutoffs
The Trail Making Test (TMT) relies on normative data to interpret performance, with standards adjusted for demographic factors such as age, education, and sometimes gender to account for their influence on completion times. Seminal normative datasets, such as those developed by Tombaugh (2004), provide stratified means and standard deviations for TMT-A and TMT-B based on a large sample of 911 healthy adults aged 18–89 years, showing that completion times increase significantly with advancing age and lower education levels. For instance, in young adults aged 20–39 years with 13–16 years of education, the mean TMT-A time is approximately 23 seconds (SD = 6.6), while TMT-B averages 51 seconds (SD = 19.3); in contrast, for those aged 70–89 years with similar education, means rise to 50 seconds (SD = 20.3) for TMT-A and 133 seconds (SD = 72.4) for TMT-B. Heaton et al. (2004) extended these with demographically adjusted norms from over 1,800 healthy individuals, incorporating regression-based corrections for age, education, and gender (with minimal gender effects observed), enabling the derivation of T-scores (mean = 50, SD = 10) that facilitate cross-test comparisons. Impairment cutoffs are typically defined using these T-scores, where scores 1–2 standard deviations below the mean (T < 40 to T < 30) indicate mild to significant deficits, particularly on TMT-B, which is more sensitive to executive dysfunction. For example, a TMT-B time exceeding 90 seconds often corresponds to a T-score below 40 in younger cohorts, signaling potential cognitive impairment, though thresholds vary by demographic adjustments to avoid overpathologizing slower but normal performances in older or less-educated individuals. These cutoffs emphasize conceptual benchmarks rather than rigid absolutes, as prolonged times (e.g., >75 seconds for TMT-A or >200 seconds for TMT-B in adults) may warrant further evaluation when adjusted for norms. Pediatric normative data differ markedly from adult standards, reflecting developmental improvements in visuomotor and executive skills, with shorter completion times overall. For children aged 9–14 years, Reitan's original child norms (updated in comprehensive batteries) report TMT-A means around 20–30 seconds and TMT-B around 40–60 seconds, decreasing with age within this range; more recent cross-cultural data, such as from a large Latin American Spanish-speaking sample of 3,337 children aged 6–17 years, confirm these patterns with education adjustments to reduce bias in low-resource settings (Ardila et al., 2017). Adjustments for low education levels across all ages, as highlighted in Heaton et al. (2004), are crucial to minimize cultural and socioeconomic confounds, ensuring equitable interpretation. Recent studies from the 2020s have expanded cross-cultural applicability, addressing limitations in predominantly Western norms. For example, demographically adjusted norms for a Scandinavian cohort aged 41–84 years (Espenes et al., 2020) align closely with Heaton's distributions but incorporate regional literacy effects, while data for native Spanish-speaking adults in the US-Mexico border region emphasize education corrections (Marquine et al., 2021). As of 2024, additional normative datasets have been published for populations including Taiwanese elderly and Korean adults, further enhancing cross-cultural utility.[25][26] These updates prioritize high-impact, diverse samples to enhance global clinical utility without altering core T-score frameworks.| Age Group | Education (Years) | TMT-A Mean (SD, seconds) | TMT-B Mean (SD, seconds) | Source |
|---|---|---|---|---|
| 20–39 | 13–16 | 23 (6.6) | 51 (19.3) | Tombaugh (2004)[27] |
| 70–89 | 13–16 | 50 (20.3) | 133 (72.4) | Tombaugh (2004)[27] |
| 9–12 | 3–6 | 28 (12) | 62 (28) | Ardila et al. (2017)[28] |