Clinical trials routinely use parent-reported assessments of children as an outcome measure [
1,
2,
3,
4]. Parents’ assessments provide additional insights into the course of a disease without imposing the extra cost associated with clinicians’ assessments. However, little data exist on the reliability of parent-report evaluations [
5]. In 2015, we published a language training app for children [
6,
7,
8,
9,
10] inviting parents to evaluate their child’s development every three months. The parents completed a Autism Treatment Evaluation Checklist (ATEC) [
11] and a Mental Synthesis Evaluation Checklist (MSEC) [
12] that assess children along five subscales: combinatorial receptive language, expressive language, sociability, sensory awareness, and health. Resultantly, more than 100,000 assessments were gathered.
The analysis of these parents’ assessments yielded several important insights into the effects of culture and physical conditions on the developmental trajectories of children with ASD. A longitudinal study investigating the impact of passive video and television watching in children with ASD (N = 3227) demonstrated that greater exposure to video and television watching was correlated with a 1.3-fold (
p = 0.0719) faster improvement in the development of expressive language, but also resulted in a 1.4-fold (
p = 0.0128) slower development of combinatorial receptive language. The differences in the sociability, sensory awareness, and health scores remained insignificant [
13]. Similarly, a prospective 3-year study looking at pretend play (N = 7069) showed that pretend play was associated with a 1.9-fold faster improvement in combinatorial receptive language (
p < 0.0001), a 1.4-fold faster improvement in expressive language (
p < 0.0001), and a 1.3-fold faster improvement in sensory awareness (
p = 0.0009); meanwhile, the effects on sociability and health were insignificant. In terms of health studies, seizures and sleep have been analyzed for their impact on development in children with ASD. An analysis of the effect of seizures (N = 8461) showed that children with no seizures improved their expressive language 1.3-times faster (
p = 0.0037), their sociability 2.3-times faster (
p = 0.0320), their sensory awareness 6.2-times faster (
p = 0.0047), and their health 20.0-times faster (
p < 0.0001), whereas the effect on receptive language was insignificant [
14]. Additionally, an investigation of the effect of sleep problems (N = 7069) showed that children with no sleep problems improved their sociability 3-times faster (
p = 0.0426) and their health significantly faster (
p < 0.0001; the exact ratio could not be calculated as the health score in children with sleep difficulties had declined relative to the baseline); the effects on receptive language, expressive language, and sensory awareness were insignificant [
15]. Finally, in a 3-year study of 6454 children, those who engaged with a specialized language therapy improved their combinatorial language scores 2.2-times faster compared to children with comparable initial evaluations (
p < 0.0001) and improved their expressive language score 1.4-times faster (
p = 0.0144). However, the differences in their sociability, sensory awareness, and health scores remained insignificant [
16].
Though these results provide interesting correlations for the impacts of multifactorial cultural and physiological conditions on ASD development, there remains resistance amongst researchers in accepting parent-reported evaluations. There is a common belief within the psychological community that parents can yield to wishful thinking, and therefore may not be reliable when assessing their own children [
17]. In order to provide clarity on the reliability of parent reports, we investigated the relationship between children’s evaluation scores and their ASD severity. The Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5), specifies three levels of ASD, depending on severity of the disorder and the support required in daily life [
18]. We hypothesized that if parents clearly understood and honestly communicated their child’s diagnosis, the reported ASD severity level would have a consistent relationship with the assessment subscales. Greater ASD severity would correspond to worse assessment scores, and vice versa. Conversely, if parents misreported their children’s diagnosis, no difference in the average assessment score would be expected between the groups.
The cross-sectional analysis of 9573 children has demonstrated statistically significant differences between levels of ASD, i.e., between mild and moderate, and moderate and severe ASD diagnosis. These differences were seen within each subscale in every age group of 3 years and older. These findings are consistent with the high reliability of parent-reported evaluations and their children’s diagnoses.