The impact of interactive visual simulations on learning statistics

Abstract

In previous studies non-interactive visual simulations in learning tasks have improved learners' conceptual understanding of statistical principles. To explore the impact of interactive visual simulations on conceptual understanding of statistical principles, an online tutorial where students could either manipulate or only observe changes of statistical graphs was developed. Overall, the tutorial supports students in learning statistical concepts immediately after working with the tutorial and two weeks after. In addition, if students could manipulate the graphs on their own, they were faster. Implications and opportunities for further investigations of interactive simulations are discussed.

Learning statistics with interactive pictures using R Shiny: Generally preferred, but not generally advantageous

“…Learning outcome was considered in terms of accurate answers and the time spent while learning [13,44,45]. Two statistical concepts, namely Cohen's d and the standard normal distribution (see Figures 1 and 2), were selected from the basic statistics course taught by the faculty of psychology at FernUniversität in Hagen in Germany [46].…”

“…They found that working with simulations enriched the classroom experience but did not improve the gain in comprehension. Iten et al [13] compared the learning outcome and learning time when using interactive simulations rather than static pictures for learning statistics. The results showed that the learning outcome was similar in both groups, but the learning time was shorter in the interactive simulation group compared with the static group.…”

“…Not many of the abovementioned studies used R Shiny to generate the simulation material. Studies have, for instance, used Excel [3], Java [40], and HTML combined with CSS and JavaScript [13]. Moreover, those studies which did in fact use the R package Shiny to generate the simulations lacked a static control condition.…”

“…Most studies on learning statistics have recommended using interactive simulations [8][9][10]. However, the findings of previous studies are inconsistent regarding the benefits of interactive simulations compared with static pictures [11][12][13][14][15]. In these studies, most of the interactive simulations were generated by other software (eg, Excel, Java, HTML, JavaScript, CSS) rather than using the R package Shiny [1,4,16].…”

Programming concepts and misconceptions in grade 5 and 6 children: Developing and testing a new assessment tool

“…In the context of programming, misconceptions can be defined as an incorrect understanding of a concept or a set of concepts that leads to making mistakes in writing or reading programs (Sorva, 2008). Research in other domains, such as physics (Edelsbrunner et al, 2018) or statistics (Iten, 2014), showed that learning about misconceptions can improve teaching a given subject. Research into misconceptions in the field of programming has mostly been focussed on adults in the past, therefore studies and assessments for children are still comparatively sparse.…”