Normative Data for the D-KEFS Tower Test in Greek Adult Population Between 20 and 85 Years Old

Abstract

Background: The Delis–Kaplan Executive Function System (D-KEFS) Tower Test (TT) is a widely used neuropsychological tool that assesses complex executive functions, including planning, cognitive flexibility, inhibition, switching, and impulsivity—key abilities often impaired in individuals with frontal dysfunction. Aims: There is currently no normative data for the D-KEFS TT in the Greek population. Consequently, it cannot be effectively used to detect executive dysfunction in neurological and psychiatric populations or for research purposes. Methods: The study sample included 249 healthy adults (28.9% male, 71.1% female) aged 20 to 85 years (M = 46.53, SD = 17.41), with educational levels ranging from secondary school graduates to master’s degree holders. Pearson correlation was used to examine the relationship between age and D-KEFS TT performance, while chi-square test assessed the effects of education and gender. Normative data were then derived from raw scores and converted into percentiles. Results: Norms were established for the following D-KEFS TT variables: Total Achievement Score, Total Rule Violations, Move Accuracy Ratio, Mean First Move Time, and Time-per-Move Ratio. Age was the strongest predictor of performance, and normative data were stratified accordingly for the Greek adult population. Conclusions: This study introduces the D-KEFS TT as a neuropsychological assessment tool for Greek adults across different age groups to evaluate complex executive functions throughout the lifespan. Unlike other D-KEFS tests, the TT had not previously been adapted for the Greek population. This study is the first to provide normative data, supporting its use in clinical practice and research.

1. Introduction

1.1. Basic Characteristics of the Delis–Kaplan Executive Function System (D-KEFS)

Executive functions, also referred to as cognitive control, are higher-order cognitive processes that support independent living, goal-setting, purposeful behavior, planning, and problem-solving skills [1]. Despite their complexity, they serve as prerequisites for other foundational cognitive skills. Executive functions are considered top-down processes, and according to Barkely (2012) [2], there is no universal definition. However, they are key functions assessed by widely used neuropsychological tests, including long-term planning, inhibitory control, cognitive flexibility, and working memory capacity, etc. [3]. Given that executive functions play a crucial role in both cognitive and physical health [4] and form the neural substrate of the frontal lobes, particularly the prefrontal regions [5], they are widely assessed in individuals with neurological and psychiatric conditions to support research, diagnosis, and treatment planning. Moreover, the frontal lobes account for approximately 40% of the human brain, making lesions and injuries in these regions impactful on a wide range of cognitive functions.

Delis–Kaplan Executive Function System (D-KEFS) is a popular neuropsychological test measuring a multiple set of executive functions, published by Delis et al. (2001) [6], which is administered in children as well as adults between 8 and 89 years old. It includes nine subtests, which are either new or revised versions of the already existing neuropsychological tests that include modifications of existing clinical tools and which involve new trial types, such as switching or alternating conditions. D-KEFS tests can be used individually or in combination to give a better evaluation of executive functions. Particularly, they consist of nine subtests: (a) Trail Making Test (TMT), (b) Verbal Fluency Test (VFT), (c) Design Fluency Test (DFT), (d) Color–Word Interference Test, (e) Tower Test (TT), (f) Twenty Questions Test, (g) Word Context Test, (h) Sorting Test, and (i) Proverb Test. Two of these D-KEFS tests have been initially designed by the D-KEFS authors. In particular, the Word Context Test was created by Edith Kaplan in the 1940s [7] and the California Card Sorting Test, which was renamed afterwards as the D-KEFS Sorting Test, was developed by Dean Delis [8]. The rest of D-KEFS tests constitute longstanding clinical executive functioning instruments [6]. It is worth noting that correlations between the D-KEFS Sorting Test, TT, VFT, and DFT were found to be “0.25 for 20- to 49-year-olds and 0.33 for those between 50–89 years” [9]. The different values per age categories suggest a modest but increasing correlation between age and test performance. Additionally, the low correlation indices provide robust evidence that the aforementioned tests do not capture the same executive functions. Till now, Greek norms of the D-KEFS were calculated for TMT and Color–Word Interference Test in adults between 20 and 49 years old [10].

Similar to previous findings, Karr et al. (2019) [11] supported a three-factor D-KEFS model, including inhibition, shifting, and fluency. However, shifting and fluency were relatively highly correlated (r = 0.696). While a bifactor structure, that is, shifting and fluency, provided a superior fit, these two factors were less consistently distinguished compared to other model structures. Hence, it can be assumed that the D-KEFS scores provide valuable information to health professionals about executive functions’ measurement, and, therefore, offer a meaningful framework for interpreting results. According to Latzman and Markon (2010) [12], the best-fitting D-KEFS model across three age bands (8–19, 20–49, and 50–89) consists of Conceptual Flexibility, Monitoring, and Inhibition, whereas MacDonald et al. (2024) [13] found that the original two-factor model, comprising Cognitive Flexibility and Abstraction, demonstrated acceptable fit in autistic adolescents and adults.

A survey conducted among members of the International Neuropsychological Society and the National Academy of Neuropsychology found that the D-KEFS battery was ranked as the third most commonly used tool for assessing executive functions [14]. The most frequently used D-KEFS tests were the Color–Word Interference (10%), VFT (8%), and TT (5%), while the DFT, TMT, and Twenty Questions Test were each used by 3% of respondents. Notably, Hacker et al. (2024) [15] found that the original versions of these tests (i.e., the Controlled Oral Word Association Test, TMT, and Stroop Interference) were administered more widely than their D-KEFS counterparts. Schmidt (2003) [16] suggests that this may be due to the lack of available psychometric data on the validity and reliability of D-KEFS tests.

1.2. Delis–Kaplan Executive Function System (D-KEFS) Tower Test (TT)

The TT falls into the category of D-KEFS tests that have been modified from existing assessments. Specifically, the Tower of London (TOL) [17,18] is one of the most widely used tests for measuring executive functions, primarily administered to younger individuals. It evaluates several executive functions, such as spatial planning, inhibition of impulsive and perseverative responses, problem solving, rule learning, and the ability to maintain a rule setting to accomplish a specific task. A similar test, the Tower of Hanoi (TOH), is also categorized as a tower test. Although the TOL and TOH share common guidelines, they do not provide equivalent feedback for assessing executive functions. Studies [19] have shown that the correlation between these two tests ranges from 0.37 to 0.61, suggesting that they measure different aspects of executive functions. Welsh, Satterlee-Cartmell, and Stine (1999) [20] reported that while these two tests are significantly correlated, their correlation is moderate. Furthermore, TOL performance can be predicted by working memory and inhibition, in contrast to the TOH.

The D-KEFS Tower Test (TT) requires various fundamental cognitive skills, including visual attention, visuoconstructional and visuospatial abilities, as well as executive functions such as spatial planning, impulsivity, inhibition, visuospatial working memory, cognitive set formation, and rule learning [6,12]. Like many neuropsychological tests, the D-KEFS TT presents increasing levels of difficulty. It is significantly correlated with the Tower of London (TOL) test, though these two assessments share only 22% of their variance. According to the authors [21], this suggests that while they measure similar executive function skills, their non-shared variance indicates they also assess distinct cognitive functions. In summary, tower tests share some overlapping variance but do not measure the same cognitive functions. Finally, the D-KEFS TT, which is primarily used with younger populations [22], represents a modified version of the TOL, specifically its latest version, but it is closer to the TOH.

The study by Trakoshis et al. (2022) [23] points out that the D-KEFS TT factorial structure includes three latent factors: impulsivity (measured by the First Move Time), behavioral disinhibition (captured by the Total Rules Violation), and difficulties or problematic planning (evaluated by the Excess Move score). Some authors emphasize the importance of assessing planning, referring to it as “the pinnacle of executive functioning” (Best et al., 2009, p. 188; [24]). Previous studies [24,25] define planning as the ability to “look ahead”, design a plan, evaluate, and monitor its execution in an efficient, organized, and strategic manner. Given that the dorsolateral prefrontal cortex (dlPFC) is activated during planning tasks such as the TOL and TOH, researchers [26,27] emphasize the significance of including tower tests in neuropsychological assessments.

The D-KEFS Tower Test (TT) consists of nine trials, which involve a set of three vertical pegs of equal length and five colored disks of varying sizes, arranged from smallest to largest. According to studies, the D-KEFS TT trials are organized into three levels of difficulty: early (trials 1 and 2), middle (trials 3 and 4), and late (trials 5–9) [28]. The results suggest that the components of each level measure different abilities. Specifically, from trial 5 to trial 9, the examinee is required to demonstrate increasingly advanced spatial planning and rule-learning skills. According to Newman (2021) [28], “the median tower trial, trial 5, is a transitional item, as it loads almost equally onto both early and late trials and is most strongly related to measures of working memory”. Initially, the participant is given a predetermined starting position for the colored disks on the three vertical pegs and is asked to arrange them in a specified ending position, as demonstrated by the examiner. The goal is to complete the task using the fewest moves possible, ranging from 1 to 26. Before starting the test, the examinee must follow two rules: only one disk may be moved at a time, and a larger disk cannot be placed on top of a smaller disk. As for D-KEFS TT scoring, various studies use different metrics, such as Total Moves, Execution Time, Total Completion Time, Total Rules Violations, Total Correct Towers, Total Achievement Score, and First Move Time [12]. According to Georgiou et al. (2017) [29], the First Move Time and Total Rules Violations are more commonly used compared to other metrics.

According to American normative data, a Move Accuracy Ratio raw score close to 1, coupled with a high Total Achievement Score, is a strong indicator that the examinee is providing careful responses and employing effective learned heuristic strategies. In contrast, if the Move Accuracy Ratio is close to 1 but the Total Achievement Score is low, the ratio may simply reflect a minimal response style. Additionally, a Move Accuracy Ratio close to 0 strongly suggests that the examinee is a minimal responder. Examinees aged 12–59 typically do not commit any Total Rules Violations, unlike younger children and older adults, who tend to commit an average of 2 or 3 violations across all items. Initial American data indicate that the Mean First Move Time for test items is 4.6 s. Faster first-time responses suggest impulsivity, while slower ones may reflect challenges with initiating activities, depression, fatigue, or insufficient effort. Furthermore, data from the entire normative sample show that the Mean Time-per-Move Ratio is approximately 3.6 s, slightly lower than the Mean First Move Average Time. Thus, it can be concluded that the D-KEFS TT provides valuable data about participants’ performance. However, it is crucial for clinicians and researchers to interpret these results carefully, as specific sub-scores are extracted in different contexts. Most D-KEFS tests utilize a game-like format without providing right/wrong feedback, which is intentionally designed to minimize unproductive discouragement and frustration caused by repeated negative feedback during test performance.

1.3. Demographics’ Effects—The Role of Gender, Age, and Education

Despite discrepancies across studies, both educational level and, more importantly, age appear to significantly impact participants’ performance on all D-KEFS tests. Regarding the D-KEFS Tower Test (TT), research highlights the significant role of age in determining scores for both young and older adult participants [21,23], making age a key predictor of performance on this test. Newman (2021) [28] emphasizes that education is positively associated with performance on later TT trials (5–9), which are considered the most challenging. Specifically, individuals with more years of education are more likely to make the required moves, as these trials demand crystallized intelligence [28], which is strongly linked to education. Gender, however, had no significant or only limited effects [30,31,32].

1.4. Study’s Scope

According to the existing literature, very few studies have used the D-KEFS TT in clinical and non-clinical populations in Greece [33,34]. These studies yielded interesting findings that highlight the test’s strong potential. Therefore, the present study aimed to calculate normative data for a typical Greek population across six age groups (20–29 years, 30–39 years, 40–49 years, 50–59 years, 60–69 years, and 70–85 years). The goal is to introduce this tool into clinical and research practice for the general population.

2. Materials and Methods

2.1. Participants

Two hundred forty nine (249) healthy adults, aged from 20 to 85 years old (28.9% male and 71.1% female), participated in the study voluntarily, anonymously, after giving their written consent. The mean age was 46.53 years (SD = 17.41), and educational levels distributed as follows: up to 9 schooling years; 10–12 schooling years; 12–16 schooling years; 16+ schooling years. To collect the sample, participants were recruited from the wider area of North Greece, specifically university students from the University of Western Macedonia and Aristotle University of Thessaloniki, as well as other adults from towns, villages, and islands of this area representatively. For middle-aged and older adult participants, aged 50 < years old, we recruited participants from the Greek Association of Alzheimer’s Disease and Related Disorders (GAADRD) and, specifically, the Day Care Centers of “Saint John” (DCCSJ) in Thessaloniki. The GAADRD provides diagnostic services, pharmacological and non-pharmacological interventions for older adults with mild-to-severe neurocognitive disorders. The sample consisted of people who had initially visited the centers for diagnostic purposes. After following all diagnostic protocols according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) [15], only Cognitively Healthy adults or those with Subjective Cognitive Impairment participated in the study.

In more detail, all participants aged 20 to 49 years met the inclusion criteria: they were native Greek speakers. The exclusion criteria included a history of addictive disorders, psychosis, or major depression; a concurrent history of neurological diseases known to affect cognitive functioning; auditory impairment that would prevent understanding of normal conversational speech; and visual acuity issues that were either non-normal or not corrected to detect visual stimuli. Participants over the age of 50 had initially visited the GAADRD for a neuropsychological and neurological evaluation due to subjective memory complaints. Specifically, these participants followed the diagnostic protocol of the GAADRD, which includes a neurological, neuropsychiatric, and neuropsychological assessment (adapted for the Greek sample from [35,36,37,38,39,40,41,42]) as well as neuroimaging (CT scan or magnetic resonance imaging) and blood tests to exclude cognitive impairment due to dementia. In addition to neuropsychological testing, psychometric tools to assess mood and behavioral issues were also administered. The official neuropsychological assessment was conducted by the neuropsychologists at the DCCSJ of the GAADRD. The assessment lasted approximately 90 min and was conducted face-to-face. The exclusion criteria for participants aged 50 and older included (a) a history of psychiatric disorders such as schizophrenia or mood disorders (major depression, generalized anxiety disorder), (b) substance abuse or alcoholism, (c) history of traumatic brain injury, (d) neurological disorders such as hydrocephalus, Parkinson’s disease, encephalitis, brain tumors, epilepsy, and stroke, (e) use of opioids, and (f) severe sensory deficits. It is worth noting that older adults with mild cardiovascular issues, such as mild hypertension, were not excluded from the study. After meeting the aforementioned criteria, participants were informed about the study’s requirements. If they expressed interest in participating, the psychologists involved in the study scheduled an appointment at the GAADRD.

2.2. Description of the D-KEFS TT Sub Scores

After reviewing the literature on the variables that can be derived from the TTs, the current study adopts the following metrics: Primary Variable—Total Achievement Score—the sum of points achieved for all items, reflecting the correct towers built within the time limit using the fewest moves. Optimal Variables: Total Rules Violation—the number of rules violated by the participant during the tower completion; Move Accuracy Ratio—the proportion of total moves made by the examinee across all items relative to the minimum required moves; Mean First Move Time—the average time taken for the first move across all items, calculated by dividing the total first-move time by the number of items administered; Time-per-Move Ratio—the average time taken by the examinee to make a move, not limited to the first one; the Mean First Move Time, Time-per-Move Ratio, and Move Accuracy Ratio—considered secondary variables—were calculated using mathematical formulas derived from primary data, which are automatically generated during the TT’s administration.

Although the Total Achievement Score is typically considered the primary variable of the D-KEFS TT, a study by Trakoshis et al. [23] found that its factorial structure comprises three latent factors—First Move Time, Rule Violations, and Excess Moves—rather than a single measure. This structure better captures the multidimensional nature of executive functions, which cannot be assessed using a single score.

In summary, the three-factor model outperformed both the unidimensional model, where all item scores loaded onto a single latent factor, and the two-factor model, which included Rule Violations and First Move Time [23]. Examining correlations among the latent factors, Trakoshis et al. [23] found that Excess Moves were negatively related to First Move Time, while Rule Violations were positively related to First Move Time. Notably, the Rule Violations factor was positively correlated with the Total Achievement Score but showed no relationship with Excess Moves, suggesting that these two factors measure distinct aspects of executive function.

2.3. Ethics

Before data collection began, consent was obtained from the Ethics Committee of the University of Western Macedonia on 21 December 2022, to approve the processing of participants’ personal data. The study was also approved by the Scientific and Ethics Committee of the GAADRD on 5 June 2024 (Scientific and Ethics Committee Approved Meeting Number: 99/5-6-24). This approval follows the General Data Protection Regulation (EU) 2016/679, enacted by the European Parliament and Council on 27 April 2016, to protect personal data. Demographic information, including age, gender, and education, was collected in accordance with EU law effective from 28 May 2018, which permits the use of sensitive personal data for research purposes. Participants were informed and consented that, upon written request, their data could be removed from the online database. The study adhered to the principles concerning personal data and the free movement of such data as outlined in the Helsinki Declaration (World Medical Association, 1997; [43]).

2.4. Procedure

Before participating in the study, individuals read an information sheet outlining that their data would be used for research purposes. They were also informed that they could withdraw at any time without consequences or the need to provide an explanation. Prior to completing the consent form, participants were assured that their records would be coded and anonymized for future research. After providing written consent, they took part in a short, structured interview to collect demographic information, including gender, age, and educational level. Each participant was assigned a unique code, combining the initials of their name with an administration series number (e.g., PM54), to maintain anonymity while allowing identification within the statistical database. D-KEFS test administration took place in a quiet environment on university premises, primarily in the morning, to minimize external interference. For older participants, data collection occurred at the GAARD in a quiet diagnostic room. All participants met the study’s inclusion and exclusion criteria and were native Greek speakers.

2.5. Statistical Analysis

Initially, parametric tests were conducted to determine whether demographics were related to D-KEFS TT performance. Specifically, a Pearson correlation test was used to examine potential relationships between the D-KEFS TT variables and age, while a chi-square test was applied to assess any gender differences in performance across the aforementioned variables and educational levels. Additionally, D-KEFS TT variables that were identified as outliers—after being transformed into z-scores (values exceeding 3)—were excluded.

Table 1. Demographic distribution of the sample.

Age Groups	20–29 n = 43	30–39 n = 58	40–49 n = 71	50–59 n = 31	60–69 n = 30	70–85 n = 31
Gender % m/f	21.2/78.8	38.6/61.4	29.6/70.4	38.7/61.3	24.1/75.9	22.6/77.4
Age (M; SD)	23.70 (2.03)	34.23 (2.47)	45.15 (3.16)	55.26 (2.3)	65.10 (3.4)	75.84 (5)
Education (range)
≤9 years				3.2%	3.4%	9.7%
10–12 years	30.2%	22.7%	27.8%	16.8%	34.5%	41.9%
13–16 years	62.8%	50%	56.9%	59.3%	55.2%	35.5%
16+ years	7.0%	27.3%	15.3%	20.7%	6.9%	12.9%

presents the demographic distribution of the sample across all age groups. Although raw scores on D-KEFS tests are typically converted to scaled scores with a mean of 10 and a standard deviation of 3, only raw scores were used in this analysis to evaluate participants’ performance. Percentile scores were also calculated to determine the threshold below which an individual’s score would fall in the lowest 10% of the population, thus categorizing it as low performance. An alpha value of 0.05 (two-tailed) was applied, and all statistical analyses were performed using SPSS software (IBM Corp., Armonk, NY, USA, Released 2020, IBM SPSS Statistics for Windows, Version 27.0).

3. Results

In accordance with the initial norms [6], the Greek DKEFS TT was age-specific but invariant across gender and education. Therefore, age-stratified norms were established for the following variables: Total Achievement Score, Total Rules Violation, Move Accuracy Ratio, Mean First Move Time, and Time-per-Move Ratio.

Table 2. Norms for the Tower Test (TT) in conditions 1 to 4 in adults aged 20–29 years.

	ΤΤ1	ΤΤ2	TT3	ΤΤ4	ΤΤ5
% percentile
95	23.00	0.00	1.02	1.40	1.86
90	22.00	0.00	1.13	1.51	1.95
75	21.00	0.00	1.25	1.87	2.11
50	20.00	0.00	1.55	2.27	2.73
25	17.00	1.00	1.78	2.63	3.17
10	16.00	1.00	2.10	4.97	3.56
M	19.23	0.39	1.56	2.61	2.69
SD	2.45	0.65	0.34	1.45	0.69

Notes: M = Mean Score, SD = Standard Deviation, TT1 = Total Achievement Score, TT2 = Total Rules Violation, ΤΤ3 = Move Accuracy Ratio, TT4 = Mean First Move Time, ΤΤ5 = Time-per-Move Ratio.

,

Table 3. Norms for the Tower Test (TT) in conditions 1 to 4 in adults aged 30–39 years.

	ΤΤ1	ΤΤ2	TT3	ΤΤ4	ΤΤ5
% percentile
95	22.00	0.00	1.23	1.11	2.44
90	20.00	0.00	1.31	1.13	2.47
75	18.00	0.00	1.43	1.55	2.66
50	17.00	1.00	1.61	3.37	3.11
25	15.00	2.00	1.87	5.62	4.24
10	14.00	2.00	2.16	6.00	4.60
M	17.00	0.95	1.68	4.22	3.39
SD	2.64	0.88	0.34	3.39	0.83

Notes: M = Mean score, SD = Standard Deviation, TT1 = Total Achievement Score, TT2 = Total Rules Violation, ΤΤ3 = Move Accuracy Ratio, TT4 = Mean First Move Time, ΤΤ5 = Time-per-Move Ratio.

,

Table 4. Norms for the Tower Test (TT) in conditions 1 to 4 in adults aged 40–49 years.

	ΤΤ1	ΤΤ2	TT3	ΤΤ4	ΤΤ5
% percentile
95	22.00	0.00	1.16	1.25	2.05
90	21.00	0.00	1.19	1.69	2.28
75	20.00	0.00	1.30	2.37	2.63
50	18.00	1.00	1.51	4.04	3.14
25	15.00	1.00	1.81	7.02	4.57
10	13/14	2.00	2.11	12.1	5.82
M	17.27	0.81	1.59	5.28	3.63
SD	2.50	0.99	0.35	3.96	1.35

Notes: M = Mean score, SD = Standard Deviation, TT1 = Total Achievement Score, TT2 = Total Rules Violation, ΤΤ3 = Move Accuracy Ratio, TT4 = Mean First Move Time, ΤΤ5 = Time-per-Move Ratio.

,

Table 5. Norms for the Tower Test (TT) in conditions 1 to 4 in adults aged 50–59 years.

	ΤΤ1	ΤΤ2	TT3	ΤΤ4	ΤΤ5
% percentile
95	22.00	0.00	1.07	1.07	2.16
90	20.00	0.00	1.08	1.16	2.65
75	19.00	0.00	1.31	1.72	2.95
50	18.00	1.00	1.57	3.18	3.50
25	15.00	2.00	1.84	7.87	4.07
10	14.00	2.00	2.06	13.3	5.70
M	17.25	0.96	1.58	5.22	3.70
SD	2.71	1.13	0.36	4.27	1.11

Notes: M = Mean score, SD = Standard Deviation, TT1 = Total Achievement Score, TT2 = Total Rules Violation, ΤΤ3 = Move Accuracy Ratio, TT4 = Mean First Move Time, ΤΤ5 = Time-per-Move Ratio.

,

Table 6. Norms for the Tower Test (TT) in conditions 1 to 4 in adults aged 60–69 years.

	ΤΤ1	ΤΤ2	TT3	ΤΤ4	ΤΤ5
% percentile
95	21.50	0.00	1.09	1.21	2.58
90	20.00	0.00	1.27	1.48	2.62
75	18.00	1.00	1.36	2.39	3.26
50	15.00	2.00	1.69	4.58	3.78
25	14.00	2.75	1.75	7.79	4.52
10	13.00	4.00	2.06	10.3	6.43
M	16.06	1.67	1.63	5.49	4.04
SD	2.85	1.27	0.33	3.65	1.14

Notes: M = Mean score, SD = Standard Deviation, TT1 = Total Achievement Score, TT2 = Total Rules Violation, ΤΤ3 = Move Accuracy Ratio, TT4 = Mean First Move Time, ΤΤ5 = Time-per-Move Ratio.

and

Table 7. Norms for the Tower Test (TT) in conditions 1 to 4 in adults aged 70–85 years.

	ΤΤ1	ΤΤ2	TT3	ΤΤ4	ΤΤ5
% percentile
95	22.40	0.00	1.14	1.42	2.98
90	19.00	0.00	1.17	1.55	3.34
75	17.00	0.00	1.22	2.57	3.81
50	15.00	1.00	1.34	3.62	4.14
25	14.00	2.00	1.68	7.16	5.31
10	12.00	2.00	1.87	10.13	7.15
M	15.45	1.00	1.42	5.1	4.69
SD	2.93	1.00	0.27	3.61	1.35

Notes: M = Mean score, SD = Standard Deviation, TT1 = Total Achievement Score, TT2 = Total Rules Violation, ΤΤ3 = Move Accuracy Ratio, TT4 = Mean First Move Time, ΤΤ5 = Time-per-Move Ratio.

provide normative classifications to guide the interpretation of the D-KEFS TT variables based on the following percentiles: Above Average (>75th percentile), Average (25th to 75th percentile), and Below Average (10th to 24th percentile). Scores above the 95th percentile were considered indicative of superior performance. Unusually Low (3rd to 9th percentile) and Extremely Low (<3rd percentile) percentiles were not calculated, as clinical groups were not included in the current study.

We first calculated the raw mean scores for each age group, along with their standard deviations. Next, we converted these raw scores into percentiles. Inferential cutoff scores were then established to identify individuals scoring in the lowest 10%, which was assumed to indicate low performance [44]. Pearson correlation analysis revealed significant associations between age and the Total Achievement Score (p < 0.001), Rules Violation (p < 0.001), Mean First Move Time (p < 0.001), and Time-per-Move Ratio (p = 0.002). In contrast, the Move Accuracy Ratio showed no association with age, and the Rules Violation score was also not significantly related to age. The significant relationship between age and D-KEFS TT scores—at least for some measures—aligns with previous findings [12,23,45]. Additionally, no significant performance differences were found between men and women on these variables. Similarly, education level did not significantly affect any D-KEFS TT variables, which is consistent with the findings from a study on Cypriot participants by Trakoshis et al. [23].

4. Discussion

In this paper, norms for the D-KEFS Tower Test (TT) have been established for the Greek adult population, with potential applications in both research and clinical settings. This effort is particularly important due to the lack of relevant studies across all age groups in this population and the necessity of standardized data for assessing the performance of Greek adults with psychiatric and neurological conditions, especially those with frontal dysfunction or traumatic brain injury [46]. While the Tower of London (TOL) has been widely used in neuropsychological assessments, interventions, and research over the past decades, recent findings [23] suggest that the D-KEFS TT provides a more comprehensive evaluation of executive functions. Unlike the TOL, which primarily captures a unidimensional aspect of executive functioning, the D-KEFS TT offers a multidimensional assessment, addressing a broader range of executive functions beyond planning.

According to the results of the current study, individuals aged 20–29 scored one standard deviation above the mean on the Total Achievement Score compared to the original normative data from an American sample collected over twenty years ago. This difference may be attributed to younger individuals’ familiarity with cognitive tasks and educational content through video games and other computerized games. Greek normative data indicate that even middle-aged individuals can achieve a satisfactory Total Achievement Score, similar to those aged 30–39. However, this score declines in individuals over 60, aligning with the findings of Batzikosta et al. [33], who reported that the Total Achievement Score can differentiate healthy older adults from those with Mild Cognitive Impairment.

Additionally, Rule Violations were more frequent, particularly among individuals over 60, consistent with the findings of Delis et al. [6] and Trakoshis et al. [23]. These studies suggest a positive correlation between age and rule violations, indicating that this measure may reflect common executive function patterns. The initial normative data from Delis et al. [6] suggest that individuals aged 20–59 typically do not violate rules, in contrast to both younger and older adults.

Regarding the Move Accuracy Ratio, and in line with the Total Achievement Score, individuals in the 20–29 age group scored approximately one standard deviation (SD) above the original norms established by Delis et al. [6]. This can be attributed to generational differences that have emerged over the two decades since the original norms were set. For the 30–59 age range, the mean scores are relatively similar. However, those in the 30–39 age group exhibited a higher Move Accuracy Ratio compared to other age groups. Consistent with the results of previous TT subtests, a decline in Move Accuracy Ratio appears to begin in those aged 60 and older. Notably, adults in the 60–69 age group scored one SD lower than the original norms.

Regarding the Mean First Move Time, it is notable that individuals aged 20–29 scored two standard deviations above the mean of the original data, consistent with the results from the previous TT subtests. Based on the mean scores, the Mean First Move Time remains stable for participants aged 40–59, while the 10th percentile is significantly lower than the original norms. Finally, participants aged 70–85 scored nearly two standard deviations above the initial American data.

Moving on to the Time-per-Move Ratio, individuals aged 20–29 now score one standard deviation higher than their American peers did 20 years ago. In conclusion, this age group has shown a significant improvement, indicating that the executive functions measured by the TT have enhanced in young people. This improvement may be attributed to factors such as greater exposure to cognitive stimulation, and increased computer usage.

It is important to note that, for young adults within the total adult normative sample, having at least one very low score is common. Low scores during testing can occur in healthy individuals for various reasons, such as lapses in attention, fluctuating effort, partial misunderstandings of task instructions, or personal weaknesses in the area being assessed. Some degree of normal intra-individual variability in executive functions may help explain the frequency of low scores within the D-KEFS normative sample. Notably, both executive functions and cognitive variability are related to the efficiency of frontal lobe functioning [47], so healthy individuals may show more variable performance on tests that assess these higher-order cognitive abilities. Therefore, it is important to emphasize that receiving one or more low scores is not unusual among healthy adults, especially if they are evaluated across many tests rather than just a few [48].

Focusing on studies examining the use of the D-KEFS TT, the literature [49] indicates that self-reported trait impulsivity, as measured by the BIS-11 scale, can predict the Total Achievement Score of the D-KEFS TT in young adults, particularly university students, after controlling for gender and alcohol use. This suggests that trait impulsivity is associated with lower executive function performance, regardless of general cognitive ability. Latzman and Markon (2010) [12] extended research on the D-KEFS factor structure by linking it to academic achievement. Their findings suggested that executive functions such as conceptual flexibility, monitoring, and inhibition predict academic success beyond the influence of general intelligence. However, a recent study by Hacker et al. (2024) [15] found that the Total Achievement Score fails to differentiate individuals with TBI from those with orthopedic injuries and controls. This may be due to the relative ease of the initial D-KEFS items (1–4), which could inflate the Total Achievement Score, masking deficits that become more apparent in more challenging tasks.

Other studies show that patients with lateral prefrontal lesions have significantly lower Total Achievement Scores on the D-KEFS TT, as well as a higher Time-per-Move Ratio and increased Total Rules Violations compared to normal controls [50]. However, Barbey et al. [51] argue that damage to the dorsolateral prefrontal cortex (dlPFC) “leads to disproportionate defects in intelligence” (p. 1367) and “does not lead to deficits in executive function”, aligning with the findings of Keifer and Tranel (2013) [52]. The authors concluded that dlPFC lesions are associated with lower general intelligence levels and that the dlPFC “is not functionally dedicated to supporting specific executive processes but may instead support higher-level mechanisms for general intelligence” (Barbey et al., 2012, p. 5) [53]. However, no significant differences were observed between groups after controlling for full-scale IQ and processing speed, supporting the notion that the observed weaknesses in patients with dlPFC lesions were not specific to executive functions. Consequently, the Total Achievement Score may not fully represent the broad spectrum of executive functions, suggesting that other latent factors or variables should also be considered. Therefore, it is crucial to analyze various TT measures to capture real differences between normal and clinical populations and to provide more representative data in future research.

Regarding the older adult population aged 65 to 92 years, studies [53,54] indicate that the D-KEFS TT, along with the TMT, predicts daily living activities. Additionally, Batzikosta et al. [33] found that the Total Achievement Score can distinguish between healthy older adults and patients diagnosed with Mild Cognitive Impairment (MCI) and is associated with sleep duration problems in older adults. Moreover, Total Rules Violation performance can differentiate between amnestic and non-amnestic MCI groups. Furthermore, Pantsiou et al. [34] demonstrated that cognitive planning can distinguish patients with early-stage vascular dementia (VaD) from individuals with vascular risk factors, identifying a specific deficit in movement precision among VaD patients. Additionally, in healthy populations, Total Completion Time has been shown to differentiate older adults with low and high levels of self-reported cognitive failures, highlighting the potential of the TT in detecting Subjective Cognitive Decline [35].

5. Conclusions

In conclusion, our findings emphasize the importance of adapting neuropsychological tests to assess executive functions in Greek adults, as data in this field remain limited. Through this study, we aim to provide researchers and clinicians with reliable and valid methods for identifying executive dysfunction and evaluating therapeutic interventions for their patients. Given the limited research on executive dysfunction in young and middle adulthood, this study highlights the need for tailored neuropsychological assessments for this population.

6. Limitations

The present study has some limitations that should be acknowledged. Although efforts were made to ensure that the sample included individuals from all educational levels, we were unable to identify participants with only a primary education, particularly those under 49 years old. This limitation may restrict the generalizability of our findings to younger individuals with lower educational levels. However, in contemporary Greece, such cases are rare, making our sample representative of the general population of young Greeks.

Additionally, while several studies have used the D-KEFS TT in various clinical populations, our study does not provide specific insights into how the current normative data compare with those studies, as no clinical populations were included. Finally, although the sample size was adequate, increasing it further could strengthen the reliability of our findings.