Closing the Gap: How an Adaptive Behavioral Based Program on a Tablet Can Help Low Performing Children Catch Up in Math: a Randomized Placebo Controlled Study
- Location: (13:026) Sydney Alrutz-salen, Blåsen hus, von Kraemers Allé 1A/1E, Uppsala
- Doctoral student: Hassler Hallstedt, Martin
- About the dissertation
- Organiser: Institutionen för psykologi
- Contact person: Hassler Hallstedt, Martin
The aim of the present thesis was to investigate the effect, using a randomized placebo controlled design, of additional adaptive, behavioral based, math training on a tablet on low performing second graders.
Early mathematic skills have a substantial impact on later school achievement. Children with poor school achievement are at risk for adverse consequences later in life. Math competencies also have consequences for the economy at large because societies are becoming increasingly dependent on skill sets including mathematics. Proficiency with basic arithmetic, also known as math fact (i.e., 3+8, 12-3), is considered to be a critical early math skill. Intervention research in mathematics have demonstrated that math fact deficits among students with low math performance can be improved with additional targeted, non- technological interventions (i.e., small-group tutoring).
The aim of the present thesis was to investigate the effect, using a randomized placebo controlled design, of additional adaptive, behavioral based, math training on a tablet on low performing second graders. The first study (study I), investigated if arithmetic skills could be assessed in a reliable and valid way on tablet. The examination showed that arithmetic scales could be transferred from paper-based tests to tablet with comparable psychometric properties, although not for a pictorial scale, and that separate norms are needed for tablet. Study II demonstrated that training on a tablet, for on average 19 hours across 20 weeks, improved basic arithmetic skills after training in the math conditions compared to control/ placebo conditions. The effects were medium sized at post assessment. There was a fadeout of effects at 6 months follow-up, where small effects were shown, and the effects decreased further at 12 months follow-up. Children with lower non-verbal IQ seemed to gain significantly more at follow- ups than children with higher non-verbal IQ. The study found no additional effects of combining working memory training and math training. Study III, using a machine learning analysis, found that children demonstrating a positive response at 6 months follow-up were characterized by having completed 90 % or more of the math program at the default level, in combination with having a fairly favorable socioeconomic background.
In summation, this work demonstrates how an adaptive behavioral based program on a tablet can help low performing children improve critical early math skills.