Evaluation of the Effectiveness of a Multifaceted Brain Training Intervention Applied in an Upper Secondary School Setting

For the last two decades different approaches to increasing cognitive ability have been investigated: Computerized working memory training, commercial video games, mindfulness and physical exercise to name a few. The present thesis concerns with cognitive training or brain training and the potential of increasing cognitive ability through targeted training interventions. The research project that formed part of this PhD was centred around a large-scale randomized controlled trial (RCT) evaluating the effectiveness of multifaceted brain training intervention promoted by the Young Brain Project. The work has resulted in three papers. The first paper presents the main results from the comprehensive RCT comparing the effect of multi-faceted brain training to potential effects obtained in an active and a passive control group. The second paper investigates individual differences in specific training gains using data from the active control group. Finally, using data from the passive control group, the third paper establishes the reliability and validity of a questionnaire on sleep quality (Pittsburgh Sleep Quality Index).

The first paper of this thesis adds to the cognitive training literature by investigating the effectiveness of a commercially developed multifaceted brain training intervention, Funktionel Hjernetræning. The intervention was delivered in a classroom-setting and can be characterised as cognitive training incorporating physical activity and social interaction. The training was delivered as part of the schedule at the upper secondary schools participating in the Young Brain project.

The research project included 17 schools, 99 classes and 2389 students. Classes were randomly assigned to one of three groups: 1) multifaceted brain training, 2) active control (dual-n-back training followed by light physical activity), 3) passive control. Training was 1 h once per week for 12 weeks. The training volume was matched between the multifaceted brain training group and the active control group. Primary outcome measures were verbal and visual working memory. Secondary outcome measures were sustained attention, psychological well-being, implicit theories about intelligence and sleep quality.

The first of the three papers presents the results of the RCT. Following a pre-registered analysis plan and applying hierarchical mixed regression models, no significant effect of multifaceted brain training was found on primary or secondary outcome measures. In addition to the main analyses, a sensitivity analysis was conducted to verify the robustness of the results. Planned sub-group analysis compared participants with high versus low intelligence, growth mindset versus fixed mindset, but showed no convincing differential effects.

The study had a number of limitations, which is addressed in the paper and in the present review. The results of the RCT are generally consistent with a large body of research on working memory training intervention showing limited effect of training interventions on activities outside training.

The second paper of this thesis explores individual differences in relation to performance change during dual-n-back training. Individual training gains vary considerably between individuals taking part in working memory training interventions. In order to be able to plan effective training interventions, it is important to understand how individual differences affect specific training outcomes. This study was conducted on a sub-group of participants from the passive control group. Specifically, participants reporting their training data on at least 9 out of 12 training sessions were included, resulting in a sample of 217 participants. Latent growth curve modelling (LGCM) was used to analyse the data. Working memory, intelligence, sustained attention, training motivation, mindset, psychological well-being, perceived stress, and sleep quality were explored as predictors of training performance. In addition, the results showed that initial training performance was predicted by working memory, intelligence, sustained attention and sleepbquality. Rate of performance change was predicted by working memory and intelligence. The results suggest that training motivation is predictive of performance midway through the training intervention, but not in the beginning of training or towards the end. Overall, the results on individual differences support a magnification effect, where individuals with higher cognitive ability start their training performing better, than individuals with lower cognitive ability. The performance of individuals with higher cognitive ability also increases at a higher rate. The implications are that for specific training gains individuals with lower cognitive abilities will not tend to catch up to individuals with high ability. The performance of individuals with lower abilities might benefit from more training time or by providing them with more effective training strategies. Moreover, the results on training motivation implicate, that actions taken to improve training motivation should be taken midway through the training intervention. Comparing the students reporting on at least 9 out of 12 training sessions with students reporting on less than 9 training sessions, we observed that participants showing higher adherence also scored significantly higher on intelligence and working memory. In addition, participants showing high adherence, had lower perceived stress and higher sleep quality, when compared to students showing low adherence. Other studies have shown a similar effect. This has implications for the implementation of working memory training intervention, given that one of the aims of working memory training is to improve cognitive deficits.

The third paper differs from the two first papers of this thesis. It investigates if a questionnaire used in the research project produces results that can be considered reliable and valid. Using data from the passive control group, the paper investigated the internal consistency, test-retest reliability and convergent reliability of the Pittsburgh Sleep Quality Index (PSQI). Low sleep quality is a common problem among young people in Denmark. Poor sleep is related to cognitive performance, psychological and somatic health. It is important to have reliable and valid measures, when screening for sleep quality. To test convergent validity the correlation between scores on PSQI and scores on psychological well-being and perceived stress was calculated. Increased sleep quality was expected to be associated with increased psychological well-being and decreased perceived stress. The Danish version of the PSQI showed satisfactory internal consistency, modest test-retest reliability and convergent validity was confirmed.

In sum, the present research project was unable to show an effect of multifaceted brain training on measures of cognitive ability and general psychological well-being. The lack of effect pertains to the specific context, where the training was applied. The results are generally in agreement with the wider literature on working memory training, showing very limited effects of working memory training interventions. Future studies might focus on the skills and strategies involved in near transfer. Considering individual differences and differentiating the training intervention according to individual ability, might be a way to improve specific training gains.