Item Response Theory Analysis of ADHD Symptoms in Children With and Without ADHD

Abstract

Item response theory (IRT) was separately applied to parent- and teacher-rated symptoms of attention-deficit/hyperactivity disorder (ADHD) from a pooled sample of 526 six- to twelve-year-old children with and without ADHD. The dimensional structure ADHD was first examined using confirmatory factor analyses, including the bifactor model. A general ADHD factor and two group factors, representing inattentive and hyperactive/impulsive dimensions, optimally fit the data. Using the graded response model, we estimated discrimination and location parameters and information functions for all 18 symptoms of ADHD. Parent- and teacher-rated symptoms demonstrated adequate discrimination and location values, although these estimates varied substantially. For parent ratings, the test information curve peaked between −2 and +2 SD, suggesting that ADHD symptoms exhibited excellent overall reliability at measuring children in the low to moderate range of the general ADHD factor, but not in the extreme ranges. Similar results emerged for teacher ratings, in which the peak range of measurement precision was from −1.40 to 1.90 SD. Several symptoms were comparatively more informative than others; for example, is often easily distracted (“Distracted”) was the most informative parent- and teacher-rated symptom across the latent trait continuum. Clinical implications for the assessment of ADHD as well as relevant considerations for future revisions to diagnostic criteria are discussed.

A data mining and item response mixture modeling method to retrospectively measure Diagnostic and Statistical Manual of Mental Disorders‐5 attention deficit hyperactivity disorder in the 1970 British Cohort Study

“…Nevertheless, the prevalence of inattentive type presentation in our sample was comparable with meta-analytic findings (Willcutt, 2012). This could be partially explained by findings from Li et al (2015), who evaluated the full scale and found that two of the items missing from our scale had significant local dependence (di5 and di7; Li et al, 2015). Also, Arias et al (2018) analyzed the full scale and found that the most information was provided by three items (dh5, di2, and di8; relative weakness of dh6, we did not consider this an aberration, because other studies using typical levels of scale measurement also found dh6 to be a weaker item in terms of information provided (Arias et al, 2018;Gomez, 2011;Li et al, 2015).…”

“…We accepted the high difficulty because it provides information at higher levels of the trait, which is desirable for our purposes. Regarding the relative weakness of dh6, we did not consider this an aberration, because other studies using typical levels of scale measurement also found dh6 to be a weaker item in terms of information provided (Arias et al, ; Gomez, ; Li et al, ).…”

“…This could be partially explained by findings from Li et al (2015), who evaluated the full scale and found that two of the items missing from our scale had significant local dependence (di5 and di7; Li et al, 2015). Also, Arias et al (2018) analyzed the full scale and found that the most information was provided by three items (dh5, di2, and di8; relative weakness of dh6, we did not consider this an aberration, because other studies using typical levels of scale measurement also found dh6 to be a weaker item in terms of information provided (Arias et al, 2018;Gomez, 2011;Li et al, 2015). Our method extends previous work that aimed to identify ADHD in BCS70 (Brassett-Grundy & Butler, 2008) by adhering more closely to the current definition of ADHD and estimating a more precise dimensional measure.…”

“…The present study adds to the literature on IRT models of ADHD, which has primarily focused on evaluating psychometric properties of items (e.g., Arias et al, 2018;Gomez et al, 2011;Li et al, 2015;Polanczyk et al, 2010). Our approach aimed additionally to minimize error and estimate a theta score as precisely as possible, through use of a large nonclinical sample and adjustment for the zero-inflated distribution of symptomatology.…”

“…IRT models have been widely recommended for measuring psychiatric and health-related constructs (Edelen & Reeve, 2007;Gorter et al, 2015;Muthen & Asparouhov, 2006;Sturm, Kuhfeld, Kasari, & Mccracken, 2017). Importantly, other authors have used IRT to evaluate psychometric item properties of DSM ADHD criteria (Arias, Esnaola, & Rodríguez-Medina, 2018;Gomez, 2007Gomez, , 2008Gomez, , 2011Gomez, , 2012Gomez, Vance, & Gomez, 2011;Li, Reise, Chronis-Tuscano, Mikami, & Lee, 2015), compare model fit in subsamples (Polanczyk et al, 2010), and provide quantitative verification of diagnosis (Lindhiem, Yu, Grasso, Kolko, & Youngstrom, 2015). These IRT studies reported good indicators of model fit in a variety of clinical and nonclinical samples.…”