Synonyms

D-KEFS

Description

The Delis-Kaplan executive function system (D-KEFS) is the first battery of tests designed exclusively for the assessment of executive functions in children and adults that has been normed on a large national sample representative of the demographic characteristics of the US population (Delis et al. 2001a). Key objectives in the development of D-KEFS measures were to provide psychologists a comprehensive tool for assessing a wide array of executive functions, including cognitive flexibility, problem-solving, conceptual reasoning, inhibition, multitasking, and nonverbal and verbal creativity.

The D-KEFS can be distinguished from most other executive-function tests by its embrace of the “cognitive process approach” in which multiple measures are generated to isolate the mechanism of a patient’s poor score on a particular task (Delis et al. 2000; Kaplan et al. 1999). With the exception of the Wisconsin Card Sorting Test (Heaton et al. 1993), the majority of existing clinical measures of higher-level cognitive functions yield a single achievement score for each task (e.g., the Category Test and the traditional Trail-Making Test). The single-score method is particularly problematic for executive-function tests because these tasks typically require both more fundamental cognitive skills (e.g., language, visuoperception, and fine motor skills) and higher-level executive functions for successful performance. The D-KEFS was developed to provide measures that assess (a) both fundamental and higher-level skills in order to evaluate if a patient’s poor performance on a task is due to an executive-function deficit or to an impairment in a more fundamental cognitive skill and (b) strategies and error types in order to more precisely characterize the nature of the executive dysfunction.

Another objective of the D-KEFS was the development of test-design features that enhanced the instruments’ sensitivity to mild executive dysfunction or mild frontal-lobe injury. For instance, switching conditions were added to several traditional executive-function tasks that previously did not require cognitive flexibility. As an example, in addition to the three standard “Stroop” conditions (color naming, word reading, and inhibition; Stroop Color Word Test, Adult), a new “Stroop” condition was developed for the D-KEFS Color-Word Interference Test that requires examinees to switch back and forth between naming the dissonant ink color and reading the dissonant word. Cognitive switching, or the ability to abandon a previous response in order to generate a novel or complimentary response (Delis et al. 2001a), is considered one of the hallmarks of executive functions and is particularly dependent on the integrity of frontal-lobe functioning. In addition, certain D-KEFS tests were designed to contain “capture stimuli” that pull for concrete or stimulus-bound responding in patients who are vulnerable to this tendency due to frontal-lobe injury.

Descriptions of the D-KEFS Tests

The D-KEFS consists of nine tests measuring a wide spectrum of verbal and nonverbal executive functions. These tests were either (a) modifications of existing clinical tests of executive functions (e.g., Stroop Color-Word Interference Test and Trail-Making Test) in order to increase their sensitivity to frontal-lobe dysfunction, (b) or modifications of tasks used in past experimental studies of executive functions but that had not been developed into standardized clinical instruments, or (c) “relatively new” tests that Delis and colleagues developed.

D-KEFS Trail-Making Test

The D-KEFS Trail-Making Test is a modification of the traditional Trail-Making Test that was first developed by Partington (Partington and Leiter 1949), modified by army psychologists, and then included in the Halstead-Reitan battery. It is one of the most commonly used neuropsychological measures for assessing executive functions. While the traditional version offers two conditions – Part A (number sequencing) and Part B (number-letter switching) – the D-KEFS version offers five conditions: visual scanning, motor speed, number sequencing, letter sequencing, and number-letter switching. The four baseline conditions enable the clinician to assess empirically if a patient’s poor performance on the switching condition is due to a deficit in cognitive flexibility or to impairment in one or more of the underlying component skills needed to perform the switching task (i.e., motor speed, visual scanning, number sequencing, or letter sequencing).

D-KEFS Verbal Fluency Test

Verbal fluency tasks are among the most commonly administered neuropsychological measures (Verbal Fluency). The D-KEFS Verbal Fluency Test is a timed measure that includes (a) a letter fluency condition that requires examinees to generate as many words as possible that start with a particular letter, (b) a category fluency condition that requires examinees to generate as many words as possible from designated semantic categories, and (c) a category switching condition that requires examinees to alternate between generating words from two different semantic categories. In general, patients with predominately frontal-lobe damage tend to have more difficulty on the letter fluency task relative to the category fluency task, whereas patients with early Alzheimer’s disease often show the opposite pattern due to a breakdown in semantic knowledge (Delis et al. 2001b).

D-KEFS Design Fluency Test

Design fluency measures, developed as a nonverbal analog of verbal fluency measures (Jones-Gotman and Milner 1977; Design Fluency), require examinees to produce as many different designs as possible that meet a particular criterion (e.g., the designs must always contain four lines) within a time interval. The D-KEFS Design Fluency Test is a modified version of traditional procedures. For each condition, the examinee is presented rows of boxes that contain an array of dots and is instructed to draw as many different designs as possible using only four straight lines. The three conditions of the test vary in difficulty, with Condition 1 (filled dots) assessing basic design fluency, Condition 2 (empty dots only) requiring examinees to inhibit connecting the filled dots while connecting only the empty dots, and Condition 3 (switching) requiring examinees to switch between connecting filled and empty dots. The switching condition, which is novel to the D-KEFS, has been found to be particularly sensitive to frontal-lobe dysfunction (e.g., Cato et al. 2004; Kramer et al. 2007).

D-KEFS Color-Word Interference Test

The D-KEFS Color-Word Interference Test is a variant of the Stroop Test (Stroop 1935; Stroop Color Word Test, Adult), a commonly used measure of response inhibition. The D-KEFS version includes two baseline conditions – naming of color squares (Condition 1) and reading of color words printed in black ink (Condition 2) – and the traditional interference condition (Condition 3) in which the examinee must inhibit reading the words in order to name the incongruent ink colors in which these words are printed in. The D-KEFS version differs from other Stroop tasks by the inclusion of a fourth condition that requires the examinee to switch back and forth between naming the dissonant ink colors and reading the conflicting words. This condition has been shown to be particularly sensitive to frontal-lobe dysfunction (e.g., Cato et al. 2004).

D-KEFS Sorting Test

The D-KEFS Sorting Test (formerly called the California Card Sorting Test; Delis et al. 1992) was designed to provide a standardized measure of conceptual-reasoning skills. Specifically, Condition 1, free sorting, requires examinees to sort cards according to eight possible target rules, including five primarily perceptual or nonverbal rules (e.g., straight versus curved outer edges), and three primarily verbal rules (e.g., clothing versus body parts). In Condition 2 (sort recognition), the examiner sorts the same sets of cards into two groups according to the eight target sorts and, after each sort, asks the examinee to identify and describe the correct rules used to generate the sort. Performance is evaluated both in terms of the total number of correct target concepts reflected in the examinee’s sorts, as well as the accuracy and level of abstraction of the examinee’s sort descriptions. This measure has been found to be sensitive to frontal-lobe dysfunction (e.g., Huey et al. 2009; Fine et al. 2009).

D-KEFS Twenty Questions Test

The D-KEFS Twenty Questions Test is a modification of a popular, informal game played by children and adults, and assesses categorical processing, hypothesis testing, and concept formation. For this task, the examinee is presented with a stimulus page depicting pictures of 30 common objects. The examinee tries to ask the fewest number of yes/no questions in order to identify the unknown target object. Multiple process measures are scored, including level of abstract thinking, error types, and response strategies (e.g., verbal versus nonverbal). This test has been found to be sensitive to focal frontal lesions (Baldo et al. 2004).

D-KEFS Word Context Test

The D-KEFS Word Context Test is an adaptation of a test developed originally for studying how children acquire word meanings (Werner and Kaplan 1952). The examinee is asked to discover the meaning of mystery words based on clues given in sentences. For each mystery word, the examinee is shown five sentences (clues) that assist with decoding the meaning of the words. This test is a measure of deductive reasoning and verbal abstract thinking and patients with focal frontal lesion exhibit deficits on this measure (Keil et al. 2005).

D-KEFS Tower Test

The D-KEFS Tower Test is a modified version of other tower tasks that have been used in experimental studies. The examinee is asked to build target towers by moving two to five disks of different sizes across three pegs in as few moves as possible while adhering to two rules: (a) move only one disk at a time and (b) never placing a larger disk over a smaller disk. The D-KEFS Tower Test assesses several aspects of executive functioning, including spatial planning, rule violations, and inhibition. This test has been found to be sensitive to focal frontal lesions (Yochim et al. 2009).

D-KEFS Proverb Test

The D-KEFS Proverb Test was modeled after a proverb interpretation measure developed in the 1950s by Gorham (1956). For the D-KEFS version, an examinee is asked to interpret proverbs in two conditions: (a) free inquiry where the examinee generates his or her own interpretations of the proverbs and (b) multiple choice where the examinee selects the best interpretation from among four alternatives. This task measures metaphorical thinking and provides a means for comparing generation versus comprehension of abstract verbal information. This test has been found to be sensitive to disorders affecting the frontal-lobes (McDonald et al. 2008; Kaiser et al. 2013).

Administration and Scoring

Instructions for administration are given in the stimulus booklet or record form. The required materials, discontinued rules, time limits, and standardized prompts are clearly displayed to the examiner in the stimulus booklet and the response form.

Each D-KEFS test is a stand-alone instrument that can be administered individually or with other D-KEFS tests. For all of the “primary” measures and many of the “optional” measures, the raw scores are converted to scaled scores, with a mean of 10 and a standard deviation of 3, for each of the following 16 age groups: 8, 9, 10, 11, 12, 13, 14, 15, 16–19, 20–29, 30–39, 40–49, 50–59, 60–69, 70–79, and 80–89. Raw scores for several of the optional measures have limited ranges in normative (nonclinical) populations, and therefore they are corrected using cumulative percentile ranks for each of the 16 age groups. The D-KEFS scoring software is available, which automatically computes many of the scoring formulas and converts raw scores into standardized scores. Most traditional executive-function tests do not provide alternate forms, which can be problematic for repeat examinations. However, the D-KEFS provides alternate forms for three of the tests that are particularly prone to practice effects (i.e., Sorting Test, Twenty Questions Test, and Verbal Fluency Test).

Normative Sample

The D-KEFS was standardized on a nationally representative, stratified sample of 1750 children, adolescents, and adults, aged 8–89 years. Stratification was based on age, sex, race/ethnicity, years of education, and geographic region. The 2000 US census figures were used as target values for the composition of the D-KEFS normative sample. There were 75–175 people in each of the 16 age groups. The D-KEFS sample was comprised of approximately equal proportions of men and women for most of the age groups; however, the older age groups had more women than men, which is consistent with the census data. The D-KEFS sample was divided into five major educational groups used by the U.S. Census: less than or equal to 8 years, 9–11 years, 12 years, 13–15 years, and greater than or equal to 16 years. For examinees between the ages of 8 and 19, the mean parental education was substituted. In addition, recent studies of the normative sample have utilized multivariate analyses to generate base rates of “low scores” on a subset of D-KEFS measures (Karr et al. 2017; Karr et al. 2018), which aid interpretation of test performance.

Historical Background

Most of the D-KEFS tests were developed in the late 1980s and early 1990s by Delis and colleagues, with the entire set of D-KEFS instruments assembled in 1994. The preliminary measures underwent a “try-out” study of approximately 300 normative subjects and 50 mixed neurological patients, and modifications were made to improve the reliability and validity of the measures.

Psychometric Data

Reliability and Validity

Evidence of the reliability and validity of the D-KEFS is discussed in detail in the technical manual (Delis et al. 2001b; also see, Delis et al. 2004; Homack et al. 2005). Internal consistency (split-half coefficients), test-retest reliability, standard errors of measurement, and confidence intervals were estimated for each D-KEFS test. The split-half reliability estimates varied across measure and across age group, and the majority of tests had moderate or better reliability estimates (Delis et al. 2001b; Homack et al. 2005). Similarly, test-retest reliability estimates varied considerably across task and age group, and most of the test-retest coefficients were adequate (Delis et al. 2001b; Homack et al. 2005). Alternate forms reliability was provided for the D-KEFS Verbal Fluency Test, Sorting Test, and Twenty Questions Test. Most of the measures possessed adequate alternate forms reliability, although the reliability coefficients for some measures from the Twenty Questions Test were relatively low (Delis et al. 2001b).

The technical manual includes evidence of validity in the form of correlations between D-KEFS measures, between the D-KEFS and other tasks (i.e., California Verbal Learning Test-II (CVLT-II); Delis et al. 2000; and the Wisconsin Card Sorting Test; Heaton et al. 1993), and findings from clinical populations (see clinical uses below). Many of the validity studies of the D-KEFS have been published in refereed scientific journals rather than in the test manual, with scientific journals often requiring more rigorous methodology than the validity studies often reported in test manuals (Delis et al. 2004).

In general, primary measures from the same test were more highly correlated, and the strength and association between variables varied considerably across the age groups. The correlations between optional (“process”) measures tended to be low, which was expected given that normative (nonclinical) populations often have reduced ranges of these scores (e.g., error measures are typically low in normative populations but significantly elevated in certain clinical populations). As evidence of convergent validity, the manual reports that the D-KEFS Sorting Test and the Wisconsin Card Sorting Test were moderately correlated in a small sample of participants (n = 23). In addition, a study by Floyd et al. (2006) demonstrated that particular clinical clusters from the Woodcock-Johnson-III Tests of Cognitive Abilities that draw upon aspects of executive functions were significantly correlated with performance on a number of D-KEFS measures. The test manual includes correlations between the CVLT-II (Delis et al. 2000), a measure of verbal learning and memory (California Verbal Learning Test), and the D-KEFS in a sample of 292 adults. Correlations between the tests were for the most part not significant, which supports the discriminate validity of the D-KEFS (for additional discussion of validity studies of the D-KEFS, see Delis et al. 2001b; Delis et al. 2004; Homack et al. 2005).

Clinical Uses

Consistent with a stated aim of the D-KEFS, numerous studies have demonstrated that the D-KEFS is sensitive to frontal-lobe dysfunction and executive-function deficits associated with a variety of neurological, developmental, and psychiatric disorders.

A series of studies by Baldo and colleagues have demonstrated that patients with focal frontal-lobe injury have deficits on a variety of D-KEFS measures, including the D-KEFS Towers Test (Yochim et al. 2009), the D-KEFS Trail-Making Test (Yochim et al. 2007), the Twenty Questions Test (Baldo et al. 2004), and the Word Context Test (Keil et al. 2005). Further evidence of the D-KEFS’ sensitivity to frontal-lobe dysfunction is found in a series of studies by McDonald and colleagues examining D-KEFS performance in patients with frontal-lobe epilepsy. Results from these studies indicate that frontal-lobe epilepsy is associated with deficits on D-KEFS measures that emphasize cognitive switching (i.e., the number-letter switching from the Trail-Making Test, design fluency switching, inhibition/switching from the Color-Word Interference Test; McDonald et al. 2005a, 2005b, 2005c) and verbal abstraction (i.e., the Proverbs Test; McDonald et al. 2008). In addition, studies utilizing quantitative magnetic resonance imaging (MRI) to elucidate the brain regions underlying performance on the D-KEFS have provided further support for the utility of particular D-KEFS measures (i.e., Design Fluency Test, Towers Test, and Sorting Test) in assessing frontal-lobe functions (see Kramer et al. 2007; Carey et al. 2008; Fine et al. 2009).

In addition, the clinical utility of various D-KEFS subtests has been demonstrated in studies of clinical populations, including multiple sclerosis (Parmenter et al. 2007), traumatic brain injury (Heled et al. 2012; Anderson et al. 2017), cardiovascular disease (Jefferson et al. 2007; Kramer et al. 2002), autistic spectrum disorders (Kleinhans et al. 2005), ADHD (Halleland et al. 2012), Parkinson’s disease (Fine et al. 2011), schizophrenia (Savla et al. 2011), and in children with heavy prenatal alcohol exposure (Mattson et al. 1999; Schonfield et al. 2001).

Cross-References