Aim and Arguments
The article by Arena and Schwartz describes the results of an experiment to combine traditional instruction and video games in the field of teaching statistics. The aim of the experiment was to infer if students who played a specially designed digital game, Stats Invaders! were able to score higher than those who used only traditional instructions. The researchers wanted to test the hypothesis if modern digital ways of instruction combined with traditional ones could outperform the latter alone in college statistics classes. They argued initially that game-based learning develops specific skills and understanding of probability distributions.
Methods
The game upon which the researchers based their experiment was a modified version of a famous game, Space Invaders! and represented a set of levels that contained different challenges. It allowed for two different gameplay modes, namely, proportion and distribution. In each stage, regardless of the mode, a player (or learner) is faced with different statistics concepts such as probability density function, mean, variance, and distribution of probabilities. The experiment involved 97 participants, half of which played or passed the game or was presented with conventional paper-based instructions for the statistics test.
Additionally, there were two categories of participants who did not participate in a game or did not read the conventional instructions. The modes were administered by the researchers as appropriate. The data on all four categories were collected prior to and after the gaming or conventional learning experience through a 10-item survey. The gathered data were analyzed using a two-way ANOVA tool.
Results
The main results showed the success of the game-based solution as opposed to conventional instruction methods. Both pre-test and post-test demonstrated a sharp increase in the rate of passage of statistics exam in those who played the game. In addition to that, the participants who successfully passed all levels were scoring higher during the written exam on probability distributions. As finishing the game granted access to written instruction afterward before the written test, the score results demonstrated an increase as compared to other forms of learning within the experiment. Proportion-Based gameplay yielded less exam passage rate than a distribution-based one.
The results were proven statistically significant in all cases except for comparison of passage and gameplay-only test results, as all P scores were less than.023. Yet, the researchers note that proportion mode delivered less stable and predictive results, which makes it slightly less productive in terms of learning performance. Also, contrary to one of the secondary hypotheses, gameplay only demonstrated similar results to a passage that were far from statistical significance.
Implications and Limitations
The relative success of the game-based learning is reported to have a broad future for utilization in teaching statistics in college or other academic settings. Importantly, the combination of learning pathways, as was initially argued, seems to demonstrate better results than if used separately from each other. Therefore, it is vital for educators to utilize video games as prominent attributes of future learning as assisting technology in teaching statistics, and possibly, other subjects as well.
The unstable correlation between gameplay-only and passage indicates that there is additional research needed to explore the limits to which this particular combination of instruction methods extends its practical utility. As for the limitations, is the loss of valuable data in the collection stage, as several game experiences failed to be accounted for, which resulted in fewer usable responses. Besides, the sample seems to be representative only of one community college, which does not allow for wide generalizability of results.