Tasks of Executive Control Peter K Isquith Robert

Tasks of Executive Control™ Peter K. Isquith, Robert M. Roth, & Gerard A. Gioia

The TEC is: Ø a standardized computer-administered measure of two essential executive functions: § Inhibitory Control § Working Memory § Based on established neuroscience methods § For ages 5 through 18 years § With 3 equivalent forms & 2 research forms § Standardized Regression Based Change Scores facilitate interpretation within and between administrations § Stable platform and accurate timing

Overview • Why the TEC? • Basics – Stimuli, Administration – Factor, Summary, & Task Scores • Psychometrics – Evidence for Reliability – Evidence for Validity • Interpretation – Cases

Rationale for the TEC Why emphasize working memory and inhibitory control? §Fundamental to enabling other executive functions §Integrity impacts social, emotional, behavioral, and academic functioning §One or both are abnormal in numerous pediatric clinical populations

Why develop a new test of working memory and inhibitory control? § Working memory assessment often limited to brief holding of simple information (e. g. , digits) § Limited evaluation of temporal or sustained aspects of working memory § Few standardized inhibitory control tasks clinically available §The most sensitive tasks for detecting working memory or inhibitory difficulties are used in neuroscience research, but not available for clinical use

Key ingredients § Manipulate working memory and inhibitory control demands in a single paradigm § Assess over a naturalistic, sustained time period § Based on empirical evidence of relationships with, and sensitivity to, brain integrity § Appropriate for use with children and adolescents § Computer assessment for standardized administration and ease of scoring

N-Back Paradigm 0 -Back (zero-back) D X Q L G T B X C T R J V N X 1 -Back (one-back) D Y N N G T B Y C B B J V Q Y 2 -Back (two-back) D Y A L A T B Y C N R N V Q Y

N-Back Paradigm § Engages working memory over an extended time period § Parametrically increases working memory load § Accuracy and speed improves with age § Poorer performance in children associated with academic, attention, and behavioral problems at school § Poorer performance in children with TBI, ADHD, others § Developed and used extensively in the context of functional neuroimaging

N-Back Paradigm f. MRI Letter n-back in sample of 24 Adolescents Hurt et al. , 2008, J Pediatrics

N-Back Paradigm Meta Analysis of N-Back in 24 Studies Owen et al. , 2005, Human Brain Mapping

Go/No-Go Paradigm

Go/No-Go Paradigm § Long history in neuropsychology and behavioral neurology § Accuracy and speed improves with age § Poorer performance in children with ADHD, TBI, others § Used extensively in the context of functional neuroimaging, including many studies of pediatric populations

Go/No-Go Paradigm: Imaging Findings Tamm et al. , 2002, J Am Acad Child Adolesc Psychiatry

The Basics Administration & Scoring

Six Tasks • • • 0 -Back / No Inhibit 0 -Back / Inhibit 1 -Back / No Inhibit 1 -Back / Inhibit 2 -Back / No Inhibit 2 -Back / Inhibit

The Basics TEC Scores

Three Types of Scores: • Factor Scores- multimodal composites across measures provide global view • Summary Scores- unimodal means across tasks provide view of each measure • Task Scores- measures of performance for each measure by task (4 or 6 tasks)

Assessing Reliable Change in Neuropsychological Function • Change as statistically significant or clinically significant (meaningful) • Clinical trials/ clinical outcomes research mostly interested in clinically meaningful change

Standardized Regression Based (SRB) Change Scores • Take into account measurement error, regression to mean, practice effects • Adjusts for demographic factors • Later scores predicted from earlier scores via regression equations: 1 B = b * 0 B + a * Age + c * Gender + Constant • Change scores standardized in TEC

Measuring Change in Response to increasing TEC Demands Novel application of SRB change metrics Compare changes in performance across increasing working memory load (i. e. , 0 -Back to 1 -Back, 1 Back to 2 -Back) within inhibitory demand conditions (i. e. , no inhibit, inhibit). How well does individual manage the increasing executive demands across tasks? May have important clinical implications for limits of their working memory and/or inhibitory control

Measuring Change Between TEC Administrations • 3 equivalent forms and 2 research forms • No differences between forms in timing, order of stimulus presentation • Equivalent stimuli substituted in each form • SRB Change Scores assist in determining between administration differences

The Details: Psychometrics

Influence of Demographics • Age accounts for the largest percentage of variance in scores (0 to 30%) • Gender accounts for <10% of variance in some scores, primarily at younger ages – Boys tend to be faster – Girls tend to be more accurate • Racial/Ethnic background accounts for <3% in any score, and <1% in most scores

Reliability • Internal consistency >. 80 for most scores – Omissions are lower, thus reported but not interpreted • Test-retest >. 70 for Factor and Summary scores – Omissions are lower, thus not interpreted

Alternate Forms Reliability • Correlations between scores for: – Form 1 to Form 2 – Form 1 to Form 3 • Factor & Summary scores between-forms consistency range. 70 -. 95 • Task scores lower, from. 50 to. 95 • Response time measures most reliable • Omissions are lowest

Evidence of Validity Based On: • • • Expected developmental changes Expected performance changes Factor Structure Relationships with rating scales Relationships with executive tasks Differences between clinical vs control groups

Evidence Based on Developmental Changes Children become: • more accurate with age • Faster with age • More consistent with age

Factor Structure Principal factor analysis (PFA) conducted as exploratory method with oblique rotational procedure (Promax). • Performed on 1176 participants (normative sample) in two age groups: age 5 -7 (n = 219; 4 subtests, 22 variables), and age 8 -18 (n = 957; 6 subtests, 39 variables). • Three factor structure for 5 -7 year olds, 4 factors for 8 -18. Ages 5 -7 Years Ages 8 -18 Years Response Control Sustained Accuracy Selective Attention Response Speed Response Variability

TEC Correlates with Scales on: • Behavior Rating Inventory of Executive Function • BRIEF-SR • Conners’ Parent Rating Scales • Child Behavior Checklist • Post-Concussion Symptom Inventory • Everyday Situations Survey • Family Assessment Device

Evidence Based on Relationships with Executive Tasks • Digit Span • Symbol Digit Memory Test • Auditory Consonant Trigrams • Reading/Math Fluency • CVLT-C

Evidence Based on Differences Between Clinical & Control Groups • ADHD-I / ADHD-C • m. TBI • LD • Fragile X

Psychometric Summary • Large age effects, small gender effects, no bias • Appropriate internal consistency and temporal stability with multiple equivalent forms • Multiple lines of evidence support validity: – Expected developmental changes – Expected performance changes – Factor Structure – Relationships with rating scales – Relationships with executive tasks – Differences between clinical & control groups – Performance improvement tracks recovery

Interpretation

Interpretive Strategy: • Factor Scores reflect overall performance across measures • Summary Scores reflect overall performance within measures • Task Scores show change with increased working memory load with/without inhibit • SRB Change Scores measure degree of change within and between administrations

Gender & Race/Ethnicity Demographics of the U. S. Population and the TEC Standardization Sample U. S. population TEC standardization samplea % % Male 51. 13 51. 68 Female 48. 87 48. 32 Caucasian 58. 46 59. 72 African American 14. 86 13. 72 Hispanic 19. 36 20. 07 7. 32 6. 49 Sociodemographic variable Gender Race/Ethnicity Other