Supplement:
Description of 18 Executive Function Tests, Grouped by Conceptual Domain
Spontaneous Flexibility and Generativity:
Alternate Uses Test (Guilford et al., 1978): In this test, the examinee is asked to provide novel, nonobvious uses for six common objects within a limited amount of time. The maximum score is 36 points.
Random Number Generation (Brugger et al., 1996; Jahanshahi et al., 2000): This task requires the examinee to produce a random sequence of 100 digits, using the response set 0 through 9. A written trial is administered, followed by an oral trial. Many measures of randomness exist for such tasks (Towse & Neil, 1998). The metric chosen for this study was Evans=s (1978) RNG, a measure of how often each response alternative follows each other response. The score ranges from 0 (each digit-pair occurring equally often) to 1 (complete predictability of pair sequences). Thus, the higher the score, the less random the sequence. The sum of the two trials was the variable included in analyses.
Tinkertoy Test (Lezak, 1982; Koss et al., 1998): The examinee is presented with 50 pieces of a standard Tinkertoy7 set and instructed to make whatever s/he wants. The creativity and complexity of the production is rated on a 12-point scale.
Inhibition of Prepotent Responses:
D-KEFS Color-Word Interference Test (Delis et al., 2001): This standardized version of the Stroop task has four conditions: 1) naming color patches, 2) reading color names, 3) naming the color ink in which contrasting color names are printed, and 4) switching between reading color names printed in contrasting color ink and naming the ink colors. Completion times are converted to age-adjusted scaled scores. The variable used in data analysis was scaled score on condition 3 (the inhibition trial).
Hayling Sentence Completion Test (Burgess & Shallice, 1997): In the critical portion of this task, the examinee is read high-cloze sentences with the last word omitted and is required to complete it with a word that is totally unrelated. Therefore, the examinee has to inhibit a strongly activated (automatic) response before generating a new response. Relatedness of the word supplied and latency of response both contribute to the raw score. Scaled score for the entire test (on a 1 to 10 scale) was used in data analysis.
Completions & Corrections Test (Manning & Brandt, 2006): This task was developed specifically for the present study. The examinee is read 12 altered idiomatic expressions and asked to repeat each one verbatim. Five of the phrases are meant to induce errors that are either completions or extensions of the original (AHip, Hip,...@), while the remaining seven are meant to induce corrections (AThe tooth, the whole tooth, and nothing but the tooth.@) Number of phrases repeated verbatim (i.e., not corrected or completed) is recorded.
Planning and Sequencing:
Porteus Maze Test (Porteus, 1965): The Vineland Revision of this paper-and-pencil maze test was administered and scored according to standard instructions. Test Age is the outcome variable.
D-KEFS Tower Test (Delis et al., 2001): In this task, the examinee attempts to build towers of varying configurations by moving wooden disks onto vertical pegs, while constrained by specific rules, as quickly as possible. Total achievement scaled score is the variable of interest.
Tic-Tac-Toe (Crowley & Siegler, 1993; Brandt, 2008): A standardized version of this well-known paper-and-pencil game was developed specifically for the present study. Sixteen trials were played, with the examiner and participant alternating who moves first. On half the trials on which the examiner started, s/he purposely made a suboptimal initial move (i.e., one other than a corner), thereby allowing the patient an advantage. The examinee is credited one point for every trial s/he wins, and is debited one point for every loss; ties result in no change in score.
Concept/Rule Learning and Set Shifting:
D-KEFS Sorting Test (Delis et al., 2001): In this task, subjects are required to sort six tokens into two groups of three according to as many different Arules@ or principles as possible. The tokens differ on many binary dimensions, both verbal and perceptual. The task requires, among other capacities, abstraction, categorization, and mental flexibility. Confirmed sorts scaled score was used in data analyses.
Brixton Spatial Anticipation Test (Burgess & Shallice, 1997): On this nonverbal, problem-solving test, the examinee must guess where, in a 10-position array, a blue circle will appear on each of 56 trials. The blue circle=s position changes from one trial to the next and is governed by a rule or pattern that changes periodically. Raw score (total number of incorrect deductions) is converted to a scaled score ranging from 1 to 10.
Verbal Concept Attainment Test (Bornstein, 1982; Bornstein & Leason, 1985): For each of 23 problem-sets, the examinee is required to select one word from each of several short lists that shares some feature with words from other lists. Number of correct problem-sets is recorded.
Decision-Making and Judgment:
Stanford-Binet Absurdities Subtest (Thorndike et al. 1986): This task requires the detection of anomalies in drawn objects and scenes. Raw number correct was the variable of interest.
Iowa Gambling Task (Bechara et al., 1998): In this computerized task, the examinee selects 100 playing cards from among 4 decks. Each selection results in a certain payout (Awin@), but some selections are also associated with losses. Unbeknownst to the examinee, selections from decks A and B tend to result in large gains but even larger losses, while selection from decks C and D result in small, steady gains. The extent to which the examinee Acatches on@ to the payoff matrix and adjusts his/her behavior accordingly is reflected in an increase in advantageous responses as the task progresses. This was operationalized as the number of C+D deck responses on the last block of 20 trials minus the number of C+D deck choices on the first block of 20 trials.
Experimental Judgment Test: This task was developed specifically for the present study. It is similar to the Cognitive Estimation Test (Axelrod & Millis, 1994; Gillespie et al., 2002), but requires social judgment instead of the estimation of quantities and magnitudes in the material world, and is less subject to the application of algorithms. The task requires the subject to estimate three attributes of the examiner, specifically his/her age, height, and weight. Score on the test is the mean percent deviation from actual values on the three items.
Working Memory and Resource-Sharing:
Trail Making Test (Reitan, 1958): This well-known test requires visual scanning, sequencing, single and dual mental tracking, and psychomotor speed. Maximum time for the simple numeric sequencing portion (part A) was set to 300 sec; maximum time for the alternating, alpha-numeric sequencing portion (part B) was set to 600 sec. The variable used for data analysis was time to complete part B minus time to complete part A.
Brief Test of Attention (Schretlen et al., 1996): In this test of auditory divided attention and working memory, the examinee must keep track of the number of letters or number of numbers s/he hears in sequences of increasing length. Number of correct responses constitutes the score.
Telephone Search/Telephone Search While Counting (Robertson et al, 1994): Components of the Test of Everyday Attention (TEA), these subtests constitute a dual-task working paradigm. They require the examinee to search a simulated telephone directory for businesses marked by special symbols while simultaneously counting tones presented via tape recording. The extent to which visual search is compromised by the simultaneous auditory task is measured by the dual task decrement score.