Collaborative Research: Investigating Gender Differences in Digital Learning Games with Educational Data Mining

尽管有证据表明，近几十年来，数学成绩的性别差异已经缩小或消失，但关于男性在数学方面优于女性的刻板印象在童年早期就出现了，并持续到成年。这些观念似乎影响了女学生对数学的兴趣和表现，以及她们对STEM职业的追求。鉴于数位学习游戏的潜在动机效益，游戏可能提供一个途径，以减少女性学生的数学焦虑，同时提高他们的自我效能和对数学的兴趣。本项目将探讨数位学习游戏是否可以减少女性学生的数学焦虑和更好的学习，而不会伤害男性学生的学习。它将通过两个现有的数字学习游戏来研究学习：Deconstruct Point，向五年级和六年级的学生教授基础数学概念（十进制数字和运算）; Angle Jungle，针对类似的年龄范围（四年级和五年级学生），并具有类似的主题设计（即，游戏地图、卡通人物），但是具有不同的游戏机制、内容（角度）和教学方法。这项研究将探讨如何以及为什么December Point在多年的几个实验过程中一直为女学生创造学习优势。在此过程中，研究人员将确定有关性别和游戏功能之间关系的原则，这些原则可以与游戏开发者分享，并用于其他游戏，从角度丛林开始。这项工作将超越传统的男女性别二元制，分析性别的多个维度，包括性别身份（例如，学生感觉自己像男孩、女孩、两者都像、都不像的程度），性别典型性（例如，有多少学生喜欢和其他女孩[男孩]做同样的事情，有多少学生觉得他们看起来像男孩[女孩]），以及性别类型的兴趣，活动和特征（例如，一个学生有多少感情或冒险的感觉）。该研究还将调查两种假设导致性别差异的途径：首先，游戏的好玩性降低了数学内容的显着性，使其不太可能暗示数学刻板印象威胁（刻板印象威胁假说）;第二，游戏的主题细节更吸引那些（更多）认同女性的学习者，使游戏更吸引他们相比，学习者谁确定（更多）作为男孩（参与假设）。在第一年，教育数据挖掘将被用来推断学生的认知和情感过程，同时玩December Point，并将数据与这两种途径预测的不同过程进行比较。在第二年，研究人员将评估假设的途径和性别差异是否在角丛林的背景下复制。在第三年，假设将通过操纵December Point在一个版本的游戏中强调数学内容，并在另一个版本中强调乐趣和好玩的功能来进一步测试。该项目将比较两个版本之间的学习结果，以更深入地探讨刻板印象威胁和参与假设。这项工作的最终目的是提供基于性别的差异，从数字游戏学习的见解，提供原则和指导，为其他研究人员和游戏设计师在开发和修改数字学习游戏。因此，该项目有可能将DecemberPoint在女孩学习成果方面的成功转移到其他数字学习游戏，并促进对性别多层面性的认识。此外，研究结果将允许调查人员修改这两个游戏，并将其提供给全国数千名小学和中学的学生。即使在这个项目期间，大约1,950名学生-包括许多来自不同人口和低数学水平的地区-将受益于December Point和Angle Jungle的学习。这个项目由NSF的EHR核心研究（ECR）计划支持。ECR计划强调基础STEM教育研究，产生该领域的基础知识。投资是在关键领域是必不可少的，广泛的和持久的：干学习和干学习环境，扩大参与干，干劳动力发展。该计划支持积累强有力的证据，为理解、建立理论解释和建议干预和创新以解决持续性问题提供信息。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Despite evidence that gender differences in math achievement have narrowed or disappeared in recent decades, stereotypes about men being better than women at math emerge early in childhood and persist through adulthood. These perceptions appear to influence female students’ interest and performance in math, as well as their pursuit of STEM careers. Given the potential motivational benefits of digital learning games, games might provide a pathway for reducing math anxiety for female students while increasing their self-efficacy and interest in math. This project will explore whether digital learning games can lead to less math anxiety and better learning in female students, while not hurting male student learning. It will study learning with two existing digital learning games: Decimal Point, which teaches foundational math concepts (decimal numbers and operations) to 5th and 6th grade students; and Angle Jungle, which targets a similar age range (4th and 5th graders) and has a similar thematic design (i.e., a game map, cartoon characters), but with different game mechanics, content (angles), and instructional approach. The study will explore how and why Decimal Point has, over the course of several experiments spanning multiple years, consistently produced a learning advantage for female students. In doing so, investigators will identify principles regarding the relationship between gender and game features that can be shared with game developers and used in other games, starting with Angle Jungle.This work will go beyond the traditional gender binary of male and female, analyzing multiple dimensions of gender, including gender identity (e.g., how much students feel like a boy, a girl, both, neither), gender typicality (e.g., How much students like to do the same things as other girls [boys], How much students feel they look like boys [girls]), and gender-typed interests, activities, and traits (e.g., how much a student feels affectionate or adventurous). The study will also investigate two pathways hypothesized to lead to gender differences: first, that the playful features of the games reduce the saliency of the math content, making it less likely to cue math stereotype threat (the stereotype threat hypothesis); and second, that the games’ thematic details are more appealing to learners who identify (more) as females, making the games more engaging for them compared to learners who identify (more) as boys (the engagement hypothesis). In Year 1, educational data mining will be used to infer students’ cognitive and affective processes while playing Decimal Point and compare data to the distinct processes predicted by these two pathways. In Year 2, investigators will assess whether the hypothesized pathways and gender differences replicate in the context of Angle Jungle. In Year 3, hypotheses will be further tested by manipulating Decimal Point’s emphasis on math content in one version of the game and enjoyment and playful features in another. The project will compare learning outcomes between the two versions to more deeply explore the stereotype threat and engagement hypotheses. The ultimate aim of this work is to provide insights into gender-based differences in learning from digital games, providing principles and guidance for other researchers and game designers in developing and revising digital learning games. Thus, the project has the potential to transfer Decimal Point’s success with girls’ learning outcomes to other digital learning games and advance knowledge on the multidimensionality of gender. Furthermore, findings will allow investigators to revise both games and make them available to thousands of late elementary and middle school students across the country. Even during this project, approximately 1,950 students—including many from districts with diverse populations and low math proficiency¬—will benefit from learning with Decimal Point and Angle Jungle.This project is supported by NSF's EHR Core Research (ECR) program. The ECR program emphasizes fundamental STEM education research that generates foundational knowledge in the field. Investments are made in critical areas that are essential, broad and enduring: STEM learning and STEM learning environments, broadening participation in STEM, and STEM workforce development. The program supports the accumulation of robust evidence to inform efforts to understand, build theory to explain, and suggest intervention and innovations to address persistent.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.