I-Corps: Interactive Statistical Decision Trees for Application in Real-world Contexts

I-Corps项目的更广泛的影响/商业潜力包括改善与培训相关的时间和成本，以准备在学术界以外的研究领域工作的数据分析师。统计方法的知识往往不能转化为应用到当前的研究问题。拟议的项目旨在减少公司和大小组织培训新数据分析师员工所需的时间。此外，该项目的目的是减少在对特定问题应用适当统计检验方面的判断错误。在商业上，该项目可以取代用来决定统计模型应用的统计参考手册库。这个I-Corps项目将统计学的学术研究应用于现实世界的研究环境，如医学研究中心的数据分析。该项目旨在帮助定量数据分析师通过一系列简单的决策步骤，使用各种情况下分析师已知的参数，从假设到统计结果的解释。原型是使用谜题/迷宫类型构建的。目前市场上可用的工具在统计测试的范围上是有限的，并且不能用假设的数据提供实际测试的演示。当前的项目原型涵盖了28种不同的参数和非参数统计测试，适用于所有四种定量数据类型：分类、顺序、间隔和比率水平数据。此外，目前的原型通过视频链接使用样本数据演示所有28个测试，提供输出结果，并演示适当的报告。该奖项反映了美国国家科学基金会的法定使命，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。

The broader impact/commercial potential of this I-Corps project involves improvement in the time and cost associated with training needed to prepare data analysts working in research fields outside of academia. Knowledge of statistical methods often does not translate to application to current research questions. The proposed project aims to reduce the time corporations and large and small organizations need in training new data analyst employees. Additionally, the project aims to reduce errors in judgment regarding the application of appropriate statistical tests to particular problems. Commercially, the project may replace libraries of statistical reference manuals utilized to make decisions about application of statistical models.This I-Corps project leverages academic research in statistics to application in real-world research contexts such as data analysis in medical research centers. The project is designed to assist quantitative data analysts in moving from hypothesis to interpretation of statistical results through a simple series of decision steps, using parameters known to analysts across a variety of contexts. The prototype was built using a puzzle/maze typology. Current market available tools are limited in scope of statistical tests, and do not provide demonstration of the test in action with hypothetical data. The current project prototype covers 28 different parametric and non-parametric statistical tests appropriate across all four quantitative data types: categorical, ordinal, interval, and ratio-level data. Additionally, the current prototype provides demonstration of all 28 tests via video-link using sample data, provide output results, and demonstration of appropriate reporting.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.