Using Digital Games with Concrete Representations To Support Attitudes Towards Mathematics

• In Event: 3-101 - Using Education Technology to Systematically Enhance Teaching and Learning Across Grades and Content Areas

Integrative Statement

Introduction
Many students approach mathematics with a negative attitude—sometimes even developing a phobia (Tobias, 1980) or fixed mindset (Dweck, 2014). Mathematics attitudes are correlated with achievement (Ma & Kishor, 1997), and may lead to detrimental downstream consequences for students’ mathematical careers (Author, 2011). Part of the difficulty in understanding mathematics is learning to manipulate abstract symbols, which poses major obstacles for learners (Author, 2011; Koedinger et al. 2007). To engage mathematical reasoning without symbolic understanding, [company] has developed digital games that utilize concrete representations that can be digitally manipulated (see Figure 1). Iterative research and development cycles between [company] and research partner WestEd over the past 2 years (Authors, 2018) have resulted in a suite of games that can support middle-schoolers’ mathematics understanding (Pope & Mangram, 2015; Authors 2015). However, the causal link between [product] and math attitudes is untested.

Hypothesis
The present study consisted of an underpowered RCT exploring whether [product]’s digital games could support middle-school students’ math attitudes. We predicted that students who play [product] games would develop more positive math attitudes than students who do not.

Study Population
Thirty-four 5th- and 6th-grade teachers from U.S. public schools were randomly assigned to conditions (student N=415).

Method
The study took place over the course of approximately 8 weeks. Teachers were assigned to either use [product] puzzles as a supplement to their mathematics instruction (at least three days a week, for at least 10 minutes each time; treatment) or conduct their mathematics lessons in the business-as-usual way (control). Students completed likert-item surveys with questions about 1) the [product] games (treatment only), 2) mathematical attitudes, consisting of a subset of 2003 PISA questionnaire (OECD, 2005), with four subscales measuring views, confidence, self-concept, and strategies towards mathematics, delivered at the beginning and end of the study. Responses were converted to a number (0 representing a neutral response) and averaged within-participants.

Results and Discussion
[product] survey results indicated that treatment students were highly motivated to continue playing puzzles: 78% indicated they would keep attempting the puzzle until they got it right and 68% indicated they would try a different way if they failed at first. Multi-level models on attitude subscale post-scores revealed the treatment evidenced positive, but non-statistical differences relative to the control group (see Table 1). However, effects were of meaningful size, particularly for views and confidence towards mathematics, where additional gains for the treatment over control were approximately 2-3.5 times larger than the control.
Differential gains were examined on the top three items that showed the largest difference between the treatment and control (see Table 2). In all three cases, gains were larger for the treatment. Generalized hierarchical models indicated that treatment students were approximately 1.5-2 times more likely to show gains relative to the control group. Taken together, results show the promise of [product] games for promoting math attitudes. More generally, results support the possibility that carefully developed digital games using concrete representations can improve student dispositions towards mathematics.

Authors

• Bryan Matlen, WestEd
  Presenting Author

• Kimkinyona Cully, WestEd
  Non-Presenting Author

• Sara Atienza, WestEd
  Non-Presenting Author

• Ryan Burke, WestEd
  Non-Presenting Author

• Randy Weiner, BrainQuake
  Non-Presenting Author

• Keith Devlin, BrainQuake
  Non-Presenting Author