Replication Kit: "Skill Models for Programming Language Concepts"

<strong>Structure</strong> <strong>data</strong>: contains the data we used for our case study <strong>skillmodels</strong>: data sets generated from the raw data in the database <strong>raw</strong>: raw data collected in SmartAPE [1] containing the source code of the students as well as the assessment results of the system <strong>results</strong>: contains the complete results of our case study Results of AUC and RMSE for each meta-parametrization and each skill model in .csv and .Rda format boxplots of our AUC and RMSE distributions for each skill model and each meta-parameter in .pdf format calculation scripts: <strong>sm_trainer.R</strong>: script to fit different models for different meta-parametrizations and test them using different performance metrics. Uses <em>data/skillmodels </em>as input <strong>comparison.R</strong>: script that performs statistical tests to compare meta-parameters. Uses <em>results//results_pfa.Rda, results//results_afm.Rda, and results/results_prop.Rda</em> as input <strong>References</strong> [1] Albrecht, Ella et al. “Experiences in Introducing Blended Learning in an Introductory Programming Course.” <em>ECSEE</em> (2018).**结构化数据**：为本案例研究所用的数据集 **skillmodels**：从数据库原始数据生成的数据集 **raw**：从SmartAPE[1]采集的原始数据，包含学生源代码与系统评估结果 **results**：包含本案例研究的全部成果： - 针对每种元参数配置与每种技能模型的受试者工作特征曲线下面积（Area Under Curve, AUC）与均方根误差（Root Mean Square Error, RMSE）结果，存储格式为.csv与.Rda - 每种技能模型与每种元参数对应的AUC与RMSE分布箱线图，存储格式为.pdf **计算脚本**： **sm_trainer.R**：用于针对不同元参数配置拟合多种模型，并通过多类性能指标开展测试的脚本，输入路径为`data/skillmodels` **comparison.R**：用于执行统计检验以对比元参数的脚本，输入路径为`results/results_pfa.Rda`、`results/results_afm.Rda`与`results/results_prop.Rda` **参考文献** [1] Albrecht, Ella 等. 《在入门级编程课程中引入混合式学习的实践经验》. ECSEE (2018).