Abstract
The informant-based rating scales are considered an important component of the functional assessment of child and adolescent behavior problems (Alberto & Troutman, 2013), and mental health and education professionals routinely utilize such scales for screening, clinical assessment, or evaluations of interventions (Holmbeck et al., 2008). In school settings, teacher reports are among the most frequently used sources of information about children’s behavior. Since teachers have the opportunity to observe their students in a range of situations on a daily basis, they can provide unique perspectives on student behavior. Although teachers have been repeatedly described as well suited for reporting on the range of normative and symptomatic behaviors observed in students (Schmitz, Saudino, Plomin, Fulker, & DeFries, 1996), there is some evidence that their ratings may be subject to bias and especially gender bias.
A meta-analytic review of studies on gender differences in ADHD concluded that teachers rated boys as significantly more impaired than girls in comparison with parental reports (Gershon, 2002). Previous research has also suggested that among children with ADHD, girls are less likely than boys to be referred for further evaluation and treatment, which at least partially may be related to the assessments made by teachers (Rucklidge, 2010). In a general population sample, both parents and teachers rated boys as having higher levels than girls of inattention and hyperactivity symptoms, but this gender difference was more pronounced for teachers, especially regarding hyperactivity/impulsivity (Ullebo, Posserud, Heiervang, Obel, & Gillberg, 2012). Sciutto, Nolfi, and Bluhm (2004) have reported similar findings concerning gender bias and hyperactivity, using a fictional written profile (a two pages long academic record, depicting a 7-year-old with significant impairments in school functioning) and presenting varying information about the child’s gender and type of symptoms (inattention, hyperactivity, or comorbid aggression) to different groups of study participants who were all teachers. The teachers were more likely to refer boys than girls with hyperactive behavior for clinical services, suggesting that differences in teacher perceptions of boys’ and girls’ behaviors may contribute to the gender differences in ADHD referrals.
Previous research, both with children (Isaksson, Nilsson, & Lindblad, 2015) and adults (Bernardi et al., 2012), has demonstrated that the ADHD diagnosis is associated with higher levels of perceived stress, that is, the subjective experience of demands that exceed their ability to cope, including feelings of pressure, perception of physiological arousal, and a sense of reduced control. In a previous study, girls reported higher levels of perceived stress than boys (Isaksson et al., 2015). It has also been reported previously that children with ADHD have lower levels of the stress hormone cortisol, a phenomenon that seems to develop and increase over time (Isaksson, Nilsson, Nyberg, Hogmark, & Lindblad, 2012). This may relate to the impaired stress tolerance reported in studies on adults with ADHD (Bernardi et al., 2012; Wender, Wolf, & Wasserstein, 2001).
Given that teachers seem to underrate hyperactivity symptoms in girls, and that symptoms of inattention do not evoke as much attention in the classroom settings as hyperactivity (Rucklidge, 2010), ADHD symptoms in girls may not be observed and correctly assessed in the school setting. If so, their chances to obtain adequate interventions early in life may be reduced. Hypothetically, this may contribute both to the higher referral age and to the higher levels of perceived stress that are characteristic of ADHD in girls: feeling not having time enough, experiencing pressure from inner demands, and feeling helpless (Isaksson et al., 2015). On the basis of this consideration, we set out to investigate potential differences in parent and teacher ratings in a clinical sample of children with ADHD, hypothesizing that (a) teachers would report lower levels of hyperactivity symptoms in girls than parents, and that (b) lower teacher ratings of ADHD symptoms would, in combination with sex, be associated with higher levels of perceived stress in the children.
Method
Procedure and Participants
A group of children and adolescents (n = 137, 107 boys/30 girls; M age 11.8 years, ±2.8) with the clinical diagnosis of ADHD according to the International Classification of Diseases (ICD-10) were recruited from child psychiatry outpatient units. The design has been described in detail in a previous publication (Isaksson et al., 2015), in which we compared this group of children with healthy comparisons in relation to perceived stress and cortisol levels. The ADHD diagnoses were based on a clinical interview, physical/neurological assessment, and neuropsychological testing by a clinical psychologist, as well as parent and teacher ratings of symptoms on the SNAP-IV (Swanson et al., 2001). Diagnostic information and the SNAP-IV ratings, collected at the time of an established clinical diagnosis of ADHD, were collected from the child’s medical record at their clinical child and adolescent psychiatry units. Information on ADHD medication was collected from parents. Considering a lower reliability of child self-report prior to age of 11 (Compas, Connor-Smith, Saltzman, Thomsen, & Wadsworth, 2001), when the scales may need simplification and rewording to adjust for cognitive ability of the child, a subset of the children/adolescents (≥11 years of age; n = 64, 50 boys/14 girls) was invited to complete the Pressure-Activation-Stress (PAS) scale (Isaksson et al., 2015). The study was approved by the regional ethical review board in Uppsala, decision no. 2009/034. Written informed consent of all participants was obtained after the nature of the diagnostic procedures had been fully explained.
Questionnaires
For parent and teacher ratings of ADHD symptoms, the SNAP-IV scale was used, including nine items for inattention and nine items for hyperactivity/impulsivity, and two control questions, scored on a 4-point scale: 0 for “not at all,” 1 for “just a little,” 2 for “quite a bit,” and 3 for “very much.” Previous studies have demonstrated acceptable internal consistency of the subscales, and a good factor structure (Bussing et al., 2008). The overall Cronbach’s alpha varied between .93 and .96 in the parent and teacher ratings.
The PAS scale focuses on two stress dimensions: “pressure” (seven items: too many things at the same time, not time enough, feel under pressure from school demands/demands at home/inner demands, feel helpless, never feel really free) and “activation” (four items: rush, eat rapidly, keep a high speed all the day, difficult to relax; Lindblad, Backman, & Akerstedt, 2008). The PAS-items are rated on a 5-point scale (0 for “never,” 1 for “rarely,” 2 for “sometimes,” 3 for “often,” and 4 for “always”). The pressure dimension has demonstrated good and the activation dimension acceptable internal consistency. The scale has been shown to discriminate between children with ADHD and healthy comparisons (Isaksson et al., 2015). The overall Cronbach’s alpha in the study group was .75.
Statistical Analysis
All analyses were performed using the Statistical Package for Social Sciences (SPSS-22), a software package commonly used for statistical analysis in several fields like health sciences. As the SNAP-IV-measured symptoms were normally distributed, comparisons between boys and girls on ADHD symptoms were made using the independent-sample t test. Gender comparisons of the prevalence of ADHD medication were made using χ2. Individual comparisons of parent and teacher ratings were made using the paired-sample t test. As the data on perceived stress ratings were skewed, we used log-transformation of the ratings in the statistical analyses. Associations between the SNAP-IV ratings and log PAS-scores were assessed using Pearson correlation as well as multiple linear regression. In the regression models, we adjusted for age, sex, and ADHD medication, as these factors may potentially impact the association between SNAP-IV ratings and ratings on perceived stress. Two-tailed tests with p values <.05 were considered significant.
Results
Mean and standard deviations of the SNAP-IV scores and the PAS scores are presented in Table 1. Parental ratings were higher for girls than for boys on inattention (t = 2.34, p = .021). No sex differences were found in teacher ratings, except for a trend toward higher ratings in boys on hyperactivity and combined symptoms (p = .085 and p = .070, respectively). There were no differences between boys and girls regarding the prevalence of having ADHD medication. When comparing parent and teacher ratings on the ADHD symptom scale in boys, no differences were found. In girls, parents rated higher than teachers on inattention (t = 3.47, p = .002, respectively), on hyperactivity (t = 3.28, p = .004), and on combined symptoms (t = 3.63, p = .002).
Parental and Teacher Ratings on the SNAP-IV Scale and the Pressure-Activation-Stress (PAS) Scale in the Study Group (n = 137).
Note. ADD = Attention Deficit Disorder; ADHD = Attention Deficit/Hyperactivity Disorder; SNAP-IV = Swanson, Nolan and Pelham ADHD symptom rating scale.
Girls reported higher levels of perceived stress on the dimension pressure (t = 2.89, p = .007). Parental ADHD symptom ratings were not associated with the stress scores. However, teacher ratings on hyperactivity, but not inattention, were negatively associated with the global score of perceived stress (r = −.39, p = .008) and with the pressure dimension (r = −.41, p = .005), indicating that children with lower teacher rated hyperactivity symptoms perceived higher levels of stress. In linear regression models, adjusted for age, sex, and ADHD medication, the association remained both for the global score (b = −.36, p = .028) and for pressure (b = −.38, p = .016). Age, sex, and ADHD medication were not associated with ratings on perceived stress in the model. Neither were there any interaction effects between sex and teacher ratings on perceived stress.
Discussion
In this clinical sample of children diagnosed with ADHD, there were differences between teacher and parent ratings for girls, but not boys, with significantly lower teacher ratings on all ADHD symptom scales. Both girls and boys with lower teacher rated hyperactivity reported higher levels of perceived stress.
Our hypothesis that teachers would give lower ratings than parents on hyperactivity symptoms in girls was supported which is in line with previous reports (Gershon, 2002; Sciutto et al., 2004; Ullebo et al., 2012). Similarly, teacher ratings on inattention were lower than parental ratings in girls. These findings may indicate a possible referral bias, where girls are less often than boys referred by teachers for further evaluation and treatment (Gershon, 2002; Rucklidge, 2010; Sciutto et al., 2004). It has been argued that the more common hyperactivity symptoms in boys, with a higher co-morbidity of disruptive behaviors (Gershon, 2002; Taurines et al., 2010), evoke more attention in the classroom situation (Rucklidge, 2010). The finding may also partially explain the previous findings of a higher boy/girl ratio in clinical samples, as compared with the adult population samples (Coles, Slavec, Bernstein, & Baroni, 2012; Nussbaum, 2012); referrals to psychiatric services during childhood are often initiated by the teachers, while such contacts are often self-initiated in adulthood. Although this study was not designed to evaluate bias in the referral process, the effect of the child’s gender on teacher ratings is notwithstanding of interest, given that the entire sample had been clinically assessed and diagnosed with ADHD/ADD, and hence the signs of clinically significant impairment were present not only at home but also in the school settings.
Hypothetically, the lower teacher ratings of ADHD symptoms may correspond to an underidentification of ADHD in girls, increasing the risk of not providing the right support or the right form of education which in turn may correspond to the higher ratings of perceived stress in girls with ADHD. Moreover, our results showed that the lower teacher ratings of hyperactivity symptoms in the presence of an ADHD diagnosis were negatively associated with higher stress ratings in our sample, despite of sex. Thus, when the hyperactivity symptoms are less prominent or underidentified in the school setting, the feeling of pressure is higher for both girls and boys. One could also speculate that these children “hold together” during the school-day, meaning that the expressions of hyperactivity are less prominent at school. This may in turn increase the feeling of pressure.
As to limitations, even though we had a considerable number of children with SNAP-IV ratings, only a subset of these children (≥age 11 years of age) also completed the PAS scale, thereby decreasing the statistical power. The analyses mainly rely on questionnaire data but the ADHD diagnoses had been validated by structured clinical investigations. For future studies, independent research observations of the child at home and at school could be compared with parent and teacher ratings to clarify the contribution of rating bias.
Footnotes
Acknowledgements
The authors are grateful to all families who volunteered and to schools and child and adolescent psychiatric units in Uppsala, Enköping, Gävle, and Falun for contributions to the recruitment procedure.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was supported by grants from Victoriafonden through the Swedish Brain Foundation (Hjärnfonden), (no specific grant no.). This research was supported by funds from the Uppsala University Hospital Research Fund (ALF) to Johan Isaksson. The research position of Frank Lindblad was financed by the Swedish Council for Working Life and Social Research, grant no. 2006-0197.