Abstract
It is estimated that around 10% of children have attentional problems. These problems are associated with poor working memory and result in children frequently making errors, becoming easily distracted and/or daydreaming, and not following instructions. These problems have a substantial negative impact on children’s learning abilities and often lead to children falling behind their peers academically, so much so that a majority of them show decreased learning abilities in reading, math, and science when compared with their typically developing peers (Alloway, Gathercole, Kirkwood, & Elliott, 2009; Gathercole & Alloway, 2006, 2008).
The 10% of children estimated by Gathercole and Alloway (2008) to have attentional problems excluded those children affected by ADHD. ADHD is estimated to have a childhood prevalence of 5% (Polanczyk & Rohde, 2007) and is one of the primary reasons for referrals to mental health services among school-age children (Barkley, 1998). These children demonstrate pervasive difficulties with inattention, hyperactivity, and impulsivity that result in a number of serious impairments in social and academic functioning (Chronis, Jones, & Raggi, 2006). These behaviors often lead to social and emotional problems expressed by disruption to peer and family relationships, high parental criticism, poor conflict resolution, specific learning difficulties, and academic failure. They may feel isolated and sad, and have low self-esteem. Up to 40% of children with ADHD also have coexisting conduct problems (Maughan, Rowe, Messer, Goodman, & Melzer, 2004) characterized by aggressive and defiant behaviors, which may further hamper their ability to succeed in school and to maintain healthy relationships.
Most often, for children who have a diagnosis of severe ADHD, the first line of treatment is stimulant medication (National Institute for Health and Clinical Excellence [NICE], 2009). Yet there is no compelling literature to suggest that stimulants are capable of improving the long-term prognosis of ADHD (Jensen et al., 2007). The NICE issued guidelines on the assessment and treatment of ADHD in 2009 and recommended that children with milder impairments receive psychological treatment as a first line treatment. In addition, it was recommended that all treatments should include psychological interventions. There is a large evidence base supporting behavior modification treatments in these children (e.g., parenting and classroom interventions), and those with moderate impairments may benefit solely from such programs offered as a first line treatment (NICE, 2009; S. Young & Amarasinghe, 2010). For those with more severe symptoms, integrated treatment packages that provide behavior modification treatments in combination with medication are more likely to be successful than “stand-alone” treatments especially in addressing comorbid problems and broad domains of impairment (Conners et al., 2001; Fabiano et al., 2009; Pelham & Fabiano, 2008; Swanson et al., 2001; S. Young & Amarasinghe, 2010).
Thus, psychological treatments are indicated for children falling within a spectrum of attention problems and who are experiencing difficulty at school academically, emotionally, behaviorally, and socially. Traditionally, the most commonly provided psychological treatments for these children are parent training programs, many of which have been devised for children with conduct problems as opposed to children with attentional problems (Sanders, Mazzucchelli & Studman, 2004; Webster-Stratton, 1981). These may, in some instances, be supplemented with individual child-centered behavioral therapy. In addition to this, school-based “classroom” and academic interventions, which involve teachers’ learning and applying behavior modification strategies, have also been found to improve children’s behavior (for a review, see S. Young & Amarasinghe, 2010). Nevertheless, it is very uncommon for group treatments to be provided directly to young children in spite of the potential for this to be a cost-effective method of delivery (NICE, 2009). Furthermore, it is unknown as to how these interventions translate into the “real-world” experiences of parents and children of the intervention content and process.
This study aimed to pilot a group intervention, the RAPID program, in a school setting, and obtain feedback on its content and process from parents/carers and children. The RAPID program was developed for children with attentional problems and bridges classical approaches by providing a direct intervention using a child-centered approach and an indirect behavioral modification intervention that is more conventionally provided through parent training and/or classroom interventions. It is a cognitive-behavioral, psychoeducational group program that teaches cognitive, social, and emotional skills and moral values to boys and girls (age 8-12). The program is designed to improve skills in attentional and emotional control, problem solving, and social skills, and thereby aims to increase academic achievement and reduce behavioral problems. It includes the RATE-C Questionnaires which can be used to evaluate outcome in the domains of attention, emotional control, social functioning, and conduct. These questionnaires were adapted from versions previously used with young adults (Gudjonsson, Sigurdsson, Eyjolfsdottir, Smari, & Young, 2009) and this study additionally examined the reliability of the revised measures for children.
There were three aims for this study. First, it was aimed to ascertain the feasibility and acceptability of conducting the RAPID program in schoolchildren. This was achieved by inviting primary school children to participate in a pilot study (irrespective of whether or not they had attentional problems), evaluating group participation by recording group drop-out rates and obtaining qualitative feedback. Second, it was aimed to examine the reliability of the RATE-C Questionnaires for self-, parent, and teacher ratings. This was tested by investigating the internal consistency of each of the scales for the three groups using Cronbach’s alpha. An acceptable level was set at .70 or above (George & Mallery, 2003). Third, we wished to determine whether the RATE-C Questionnaires were sensitive to change over time, that is, whether they can be used as outcome measures by comparing pre- and postratings completed by parents/carers.
Method
Participants
Participants were male and female pupils attending two mainstream primary schools in the South West London area, aged 8 to 11 years, in Years 4, 5, and 6 (N = 433). A total of 88 parents (20.3% response rate), their children (47 males, 53.4%; 41 females, 46.6%) and teachers (n = 15) consented to participate, of whom 48 children and parents/carers consented to participate in the group treatment (28 males, 58%; 20 females, 42%).
Measures
RATE-C Questionnaires
These are 32-item child questionnaires that record self-report (RATE-CS) and informant report (RATE-CI) of current functioning on a 5-point Likert scale for the past month (never to almost all the time). The RATE-CI is a parallel version of the RATE-CS. The possible score for each scale ranges between 8 and 40, with the possible total score ranging between 32 and 160. The questionnaires provide a Total and four subscales: (a) Attention scale—items related to inattention and impulsivity; (b) Emotional scale—items related to anxiety, dysphoria, anger, and frustration; (c) Conduct scale—items related to verbal and physical aggression, damage to property, theft, and risk-taking behavior; and (d) Social scale—items related to the ability to positively engage and socialize with others. The RATE-C Questionnaires were adapted for use with children from the original RATE Questionnaires (Young & Ross, 2007) which have good reliability and validity (Gudjonsson et al., 2009; Gudjonsson, Sigurdsson, Adalssteinsson, & Young, 2012; Gudjonsson, Sigurdsson, Gudmundsdottir, & Sigurjonsdottir, 2010). In the present study, parents/carers and teachers completed the RATE-CI.
Acceptance and feasibility was assessed by group drop-out rates and a semistructured interview administered independently to children and parents who had participated in the group condition; 21 children and 16 parents/carers participated in the interview. The interview questioned their views on the program content, the materials used to support treatment (i.e., the children’s Workbook and the Coach Guide), the program process (e.g., duration of the program, the model of delivery that integrated group and individual coaching), and their experiences of transferring skills to daily life. Parents/carers were additionally asked about their experience of the parent training and role as a “coach.” Participant responses to questions were quantified on a metric related to each topic and analyzed descriptively.
Intervention
The RAPID program (Young, 2009) is a cognitive-behavioral, structured manualized group intervention that has a detective theme. The aims of the program are as follows:
Provide skills to children who are inattentive and/or present behavioral problems at home or at school.
Provide training and skills to those involved in the child’s care (e.g., parents and teachers) to better support and work with these children.
Provide a structured process to support parents to teach their children to transfer skills learned in the program to their everyday lives.
Provide a structured protocol that encourages collaboration between agencies involved in the child’s care (e.g., between health and/or education and parents).
The objectives of the program are to teach children and those involved in their care psychological techniques to improve attention span, develop better self-control, and prosocial competence. This is achieved through a curriculum of cognitive, behavioral, and emotional skills, and social values that teach attention and listening skills, impulse control, emotional control, problem-solving skills, interpersonal skills, and social-perspective taking. Children are taught self-regulation skills and an awareness of how their thinking and emotions affect their behavior. They are taught critical reasoning skills and alternative and consequential thinking. The curriculum is designed to help children develop the attitudes and values required for prosocial competence and to provide a forum to rehearse the social skills and values that are required for the development of prosocial competence. The nine sessions each last an hour and are delivered by teaching “Thinking Tools” to cue specific processes (e.g., to stop and think, to evaluate information), strategies (e.g., to control anxiety or anger), and problem-solving skills. These skills are rehearsed by engaging the children in different activities (e.g., games, role-plays, exercises) and through the use of a workbook kept by the children.
The program provides a direct intervention using a child-centered approach and this is complimented by an indirect intervention via parents who are trained to “coach” their child. The aim of the coaching role is to practice and reinforce the “Thinking Tools,” skills and strategies learned in the sessions, and help the child transfer these skills to everyday use. This is achieved by dedicated individual meetings between child and parent/carer (or another appropriate person) which are arranged at home between each group session. During these meetings, the child revises group sessions using his or her workbook and completes homework assignments. The parents/carers receive training for their “coaching” role, including psychoeducation and behavioral modification techniques in three 2-hr sessions held at the beginning, middle, and toward the end of the group program. Parents/carers are additionally supported by the written material in a Coach Guide which provides specific guidance and advice on how to structure the coach–child meetings. It includes information about the topics covered in each session, outlines behavioral strategies, and provides a guide to stimulate discussion and reinforce group learning.
Procedure
Letters containing information about the study were distributed to parents by the schools. These emphasized that the study was being run independently, teachers would not have access to their individual data, and nonparticipation would not in any way affect their school progress. The researchers attended the school at the end of the school day for a 1-week period to meet with parents and answer any questions. Parents also had the opportunity to ask questions by contacting the researchers directly by telephone and/or discussing participation with form teachers. A written consent for participation was required for both parent/carer and child. Following parent/carer consent, a children’s version using age-appropriate language was given to parents to discuss with their child. Once child consent was received, the researchers also met with the children in school to ensure their understanding of the consent. A letter and a consent form were subsequently provided to class teachers using the same method as for parents/carers. Following their consent, pupils in Years 4, 5, and 6 (age 8-11; n = 88) completed the RATE-CS in class (they were assisted by the researchers). The RATE-CI was sent out to parents and distributed to teachers in school. The completed questionnaires were returned in sealed envelopes.
For participants in the group intervention, separate information and consent forms were circulated using the same methodology. Forty-eight parents/carers and their children consented to participate: 27 from School 1 and 21 from School 2. Six groups were run, once per term, during the academic year as an after-school activity. At School 1, 11 children were in Group 1, 8 in Group 2, and 6 in Group 3. At School 2, 10 children were in Group 1, 6 in Group 2, and 5 in Group 3. In School 1, the groups were run by a clinical psychologist and/or assistant psychologists; in School 2, it was run by teachers (one of whom was the Special Educational Needs Coordinator). All those delivering the program attended a 3-day comprehensive training program. On completion of each group and using the same methodology described earlier, the RATE-C Questionnaires were sent to parents/carers to assess outcome. Parents and children attending School 1 were invited to participate (independently) in a postgroup interview to provide feedback on their views of the process and content of the group. Children participated in interviews in school at the end of the day. Parent interviews were held at the convenience of parents, either at school at the end of the day, by home visits, or by telephone.
Results
Acceptability and Feasibility of the RAPID Program
Four children dropped out of the group (all within the first three sessions) because they no longer wanted to attend, leaving 44 (92%) group completers. The mean age of children completing the group was 8.93 (SD = 1.00; range = 8-11 years). A total of 54.5% were boys, 3 were diagnosed with ADHD, 2 of whom were taking stimulant medication. Eight children (18.2%) had special educational needs (2 from School 1 and 6 from School 2). Twenty-six children attended all nine sessions (59.1%), 12 attended eight sessions (27.3%), 3 attended seven sessions (6.8%), 2 attended six sessions (4.5%), and 1 child attended five sessions (2.3%). All three coach training sessions were attended by 15 parents/carers (34.1%), 7 attended two sessions (15.9%), and 12 attended one session (27.3%). Ten parents/carers did not attend any coach sessions, and facilitators contacted these parents/carers by telephone to discuss the coaching role and gave instructions on how to use the Coach Guide.
Parent interviews
Postgroup interviews were completed by 21 children (84%) and 16 parents (64%). All the parents viewed the program positively stating that the police-detective theme was an engaging feature of the program and that it taught children important and relevant skills, for example, emotional control techniques, understanding body language, and problem solving. The most challenging aspect of the program was encouraging the child to complete homework assignments (63%).
The coach training was perceived to have been very helpful, especially sharing the experience of fulfilling the coach role with other parents (79%). In preparation for coach sessions, parents regularly referred to the Coach Guide, which they found to be a helpful resource (75%). Most parents ran coach sessions with children between sessions (80%), and some commented that appointing dedicated time with their child in this structured activity had been unexpectedly rewarding (38%). All parents stated that they had gained knowledge from the program that would be of future benefit, including how to control their own emotions, the importance of taking time out to talk to their child about their problems, and learning how best they can support their child effectively to resolve those problems.
Nearly all parents (94%) stated that they had observed postgroup improvements in their child’s functioning. One third of parents stated that, in general, they believed that their relationship with their child had improved due to participation in the program (31%). The most frequently parent-rated improvements were emotional control (44%), attentional control (31%), and problem solving (19%). Others included confidence, self-esteem, and perspective-taking.
Child interviews
Most children thought the sessions were the right length (81%); three would have liked them to have been longer and one child shorter. All the children were very positive about the program and stated that they most enjoyed the activities, games, and role-plays (76%) and/or applying the “Thinking Tools” (56%). They rated homework assignments (38%) to be the least popular aspect of the program.
The children were provided with a handbook, the “RAP book,” which included exercises and summaries of each session. This became a personalized resource which they decorated in keeping with session themes. All the children stated that they liked their RAP book because it encouraged them to be creative (81%) and because it was a useful revision resource (76%).
The majority of the children said they met with their coach regularly (81%), two children had coach sessions on alternate weeks, and two rarely met with their coach. They said that the two most helpful aspects of the coaching sessions were to complete homework tasks (62%) and/or to consolidate new skills (43%).
All the children endorsed postgroup improvements and most children (90%) perceived these improvements to have impacted positively on their interpersonal relationships, both with their family members and/or with their peers. Some children also noted that their behavior and concentration had improved in class (29%). Only two children stated that they perceived no self-improvement from attending the program. The most frequently child-rated improvements were emotional control (67%), behavioral control (57%), problem solving (52%), interpersonal relationships (38%), and attentional control (24%).
The RATE-C Questionnaires
Table 1 shows the means, standard deviations, and reliability of the RATE-C Scales for boys and girls, obtained from self-, parent/carer, and teacher ratings.
Means, Standard Deviations and Reliability of RATE-CS and RATE-CI Scales.
Note: RATE-CS = RATE Child Questionnaire-Self-Report; RATE-CI = RATE Child Questionnaire-Informant Report.
Reliability of the RATE-C Scales
To investigate the internal consistency of the RATE-CS and RATE-CI Questionnaires, Cronbach’s alpha was computed for the four subscales and the Total score of the RATE-C Scales. The reliability of the Total scores was excellent. The reliability was “satisfactory” to “good” for most of the subscales with the exception of the children’s self-reported social scale. Two of the parent-carer-rated subscales for girls (conduct and social) were borderline. Overall, the reliability of the parents’/carers’ and teachers’ scales were consistently higher than those of the child.
To investigate the extent to which the three ratings were related, these were correlated by using Pearson correlation coefficients (see Table 2). Parent/carer and teacher ratings significantly correlated on all scales and the Total score. Similarly, child ratings significantly correlated with parents/carers with the exception of the Social scale. Child ratings significantly correlated with those of teachers on the Social scale but were low on the Conduct scale and the Total score. Thus, there was good agreement between teachers’ and parents’/carers’ reports on all the ratings and satisfactory agreement between the children and their parents/carers. By comparison, agreement between children and their teachers was much weaker.
Correlation Between the RATE-CS and RATE-CI Scales.
Note: RATE-CS = RATE Child Questionnaire-Self-Report; RATE-CI = RATE Child Questionnaire-Informant Report.
p < .05, two-tailed. **p < .01, two-tailed.
Gender differences
Multivariate analyses were performed on the dependent measures (ADHD, Emotion, Conduct, and Social scales) with sex (boys vs. girls) as the fixed factor, and age (continuous variable) as a covariate. For teacher ratings, there was a significant main effect between boys and girls, Pillai’s Trace = 0.124; F(4, 72) = 2.54, p < .05; η2 = .124, but no significant effect for age, Pillai’s Trace = 0.121; F(4, 72) = 2.48, ns; η2 = .121. When the results for the teacher-rated dependent variables were considered separately, the only difference to reach statistical significance using a Bonferroni adjusted alpha level of .0125 was the Conduct scale, F(1, 75) = 10.07, p = .002, η2 = .118.
For child ratings, there was a significant main effect between males and females, Pillai’s Trace = 0.133; F(4, 82) = 3.15, p < .05; η2 = .133, but no significant age effect, Pillai’s Trace = 0.018; F(4, 82) = 0.38, ns; η2 = .018. When the results for the child-rated dependent variables were considered separately, the only difference to reach statistical significance using a Bonferroni adjusted alpha level of .0125 was again the Conduct scale, F(1, 85) = 8.82, p = .004, η2 = .094.
For parent ratings, there was no significant main effect between males and females, Pillai’s Trace = 0.32; F(4, 79) = 0.65, ns; η2 = .032, or age, Pillai’s Trace = 0.063; F(4, 79) = 1.33, ns; η2 = .063.
Sensitivity of RATE Scales to change over time
Paired-samples t tests were conducted to examine whether the RATE-C Scales were useful outcome measures. Table 3 shows the results indicating that parents/carers rated significant improvement on all the Scales apart from the Social scale with medium to large effect size.
Pre- and Posttreatment Scores on the RATE-CI Scales for Group Completers.
Note: RATE-CI = RATE Child Questionnaire-Informant Report.
p < .05, two-tailed. **p < .01, two-tailed.
Discussion
This study aimed to determine the acceptability and feasibility of delivering the RAPID group intervention to primary school children, to examine the reliability of the RATE-C Questionnaires, and to determine whether these Questionnaires are sensitive to change over time. The program was successfully run by health care professionals and teachers trained to run the program and its feasibility for delivery in a school setting was demonstrated by the very low drop-out rate. Acceptability was indicated by qualitative feedback from parents/carers and children that the program was enjoyed and valued. Analysis of the outcome interviews clearly supported the delivery of a direct intervention with young children. Both parents and children viewed the program positively, stating that they had gained knowledge and skills. Nearly all (94%) parents/carers said that they observed postgroup improvement in their child’s functioning, and one third stated that their relationship had improved overall with the child. All the children endorsed postgroup improvements and 90% perceived these to have impacted positively on their family and/or peer relationships.
A novel aspect of the program was the inclusion of the coach role, which integrates a direct child-centered treatment with an indirect intervention via parents. The primary aim of the coach role was to aid the transfer of acquired skills within the group into daily activities, and this role may be fulfilled by parents/carers, family members, or friends. When fathers participate, programmes have been shown to be more effective (Fabiano, 2007). In the present study, all the coaches were parents/carers. Parents endorsed the “coaching” parent training aspect of the program and used the handbook as a resource to facilitate their coaching sessions (i.e., the Coach Guide). Many parents commented that this structured activity had been unexpectedly rewarding. The “coaching” paradigm used by the program introduces a collaborative partnership between parent/carer and child and this aspect, and its terminology may be more acceptable to parents who perceive this as being less pejorative than “parent training.” It may also avoid any perception that the onus of blame for their child’s behavior lies with the parent/carer, for example, that they are “bad” parents who need to receive training in parenting skills. This aspect of the program additionally provides a structure for coaching sessions which may also be helpful for parents/carers who lack organization skills themselves and/or who may have ADHD (Harvey, Danforh, McKee, Ulaszek, & Friedman, 2003).
Reliability of the Total score on the RATE-C Scales was excellent for both males and females (i.e., Cronbach’s α greater than .8; George & Mallery, 2003). Individual subscales showed some variability but generally reasonable reliability; the one exception being the child-rated Social Scale. This finding replicates the previous finding of weakness in the adult version of this scale (Young, 2004), nevertheless, the Social Scale has been shown to be a good predictor of life satisfaction among college students (Gudjonsson et al., 2009). The RATE-C Scales were adapted from adult versions for use in children and the findings suggest that they are appropriate instruments to evaluate functioning in children. Comparison of pre- and postgroup parent/carer ratings further indicated that they can be used to measure treatment outcomes. In the present study, when using the RATE-C measures as indicators of outcome, the RAPID intervention was found to be effective in improving attention, emotional control, and conduct with medium to large effect sizes. The exception was the Social scale which was the weakest scale with respect to reliability.
These findings should be interpreted in light of study limitations. First, this study was conducted with ordinary schoolchildren (i.e., they were not identified as having any problems that require intervention) and to demonstrate the effectiveness of the RAPID program, studies need to include a relevant sample and compare them with a control group. Nevertheless, this highlights a strength of the program as it was effective in “normal” schoolchildren. Hence it is anticipated that the program would have greater benefits for children with impairments in attention, emotional control, and/or conduct, as floor/ceiling effects would be avoided. Second, outcome data were collected by the group facilitators, and this may have led to a positive response bias. Future research should evaluate the program on children referred to health services, using a randomized controlled trial (RCT) design, and include a follow-up period.
A strength of the RAPID intervention is that it meets the broad needs of children with attention problems. Gathercole and Alloway (2008) identified 10% of schoolchildren with attentional problems leading to impairment. This excluded the 5% of children with ADHD (Polanczyk & Rohde, 2007). Thus, 15% of schoolchildren may benefit from this intervention. Those with transitional problems of adjustment (e.g., unsettled due to family circumstances) may also benefit from this approach.
Footnotes
Acknowledgements
I am grateful to the two schools for supporting the project and for the participation of the children, parents/carers, and teachers. I am also grateful to the assistant psychologists who assisted with the delivery of the RAPID program, data collection, and scoring; in particular, Lucy Burt, Leah Dawson, Amie Doidge, Gareth Hopkin, Emily Goodwin, and Zoe Huntley.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Susan Young is author of the RAPID intervention.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
