Student Grade Data: Grit

Estimating Student Fixed Effects in Explaining Grade Outcomes: The Importance of Grit in Determining Student Performance

Grit is a term that has emerged in the literature as a way of describing a person’s persistence over time to overcome challenges and accomplish goals. Past measures of grit have been based on self-reported data. We develop a model of student grit that does not rely on self-reported data by employing a rich data set and fixed effects regression techniques to model student letter grade in a course, a proxy for student learning and success, while controlling for the student’s measured cognitive performance, sociodemographic variables, academic rank, relative high school performance, subject matter of the course, instructor, advisor and an individual student effect, interpreted as grit. We find the individual student effect explains significant variation in student academic performance in the classroom. We also demonstrate a method for predicting student grit using secondary education data.

Data represents a cohort of economics majors at a public four-year, full-time selective university located in the midwest. Rows represent a single university course taken by the student.

Columns:
Grade Grade earned by the student
Parent_Work_Income Parent's income, normalized
SCHOOL_GPA Student's high school grade point average, normalized
ACT_COMPOSITE Student's composite ACT score, normalized
STUDENT_RANK Student's college rank when the course was taken (Freshman, Sophomore, etc.)
Age Student's age when the course was taken, normalized
GENDER_DESC Student's self-identified gender,
URM Under-represented minority status
Student.22 Student indicators
Student.01
Student.36
Student.04
Student.05
Student.23
Student.15
Student.28
Student.27
Student.10
Student.11
Student.12
Student.13
Student.14
Student.30
Student.16
Student.18
Student.31
Student.19
Student.20
Student.03
Student.09
Student.26
Student.21
Student.02
Student.06
Student.32
Student.08
Student.29
Student.07
Student.17
Student.25
Student.33
Student.34
Student.24
Student.35
SubjECON Subject indicators
SubjMGT
SubjFIN
SubjENGL
SubjACCT
SubjCIS
SubjMKT
SubjCOMM
SubjMATH
SubjBIOL
SubjBLAW
SubjHIST
SubjAE
SubjPSY
SubjMUS
SubjELCT
SubjPE
SubjART
SubjPOLS
SubjSPAN
Instructor.04 Instructor indicators
Instructor.08
Instructor.07
Instructor.03
Instructor.16
Instructor.05
Instructor.12
Instructor.02
Instructor.06
Instructor.11
Instructor.13
Instructor.10
Instructor.01
Instructor.14
Instructor.09
Instructor.15
Advisor.02 Advisor indicators
Advisor.01
Advisor.04
Advisor.11
Advisor.05
Advisor.09
Advisor.07
Advisor.03
Advisor.06
Advisor.10
Advisor.08
Advisor.12

在阐释学业成果的固定效应估计中：毅力在决定学生表现中的重要性

毅力一词在文献中作为一种描述个人随时间克服挑战并实现目标持续性的概念而崭露头角。过往对毅力的衡量多基于自我报告的数据。本研究通过运用丰富数据集及固定效应回归技术，构建了一种不依赖自我报告数据的_student_毅力模型，用以模拟学生在课程中的字母等级成绩，该成绩可作为学生学习与成功的代理指标，同时控制学生的测量认知表现、社会经济变量、学术排名、相对高中表现、课程内容、教师、导师以及个体学生效应（视为毅力）。研究发现，个体学生效应可以解释课堂中学生学术表现的显著差异。此外，我们还展示了利用中等教育数据预测学生毅力的方法。

数据代表了一家位于中西部地区的公立四年制全日制选择性大学经济学专业学生的群体。行代表学生所修的单门大学课程。

列：
成绩 - 学生获得的等级
Parent_Work_Income - 父母收入，已归一化
SCHOOL_GPA - 学生的高中平均成绩，已归一化
ACT_COMPOSITE - 学生的ACT综合分数，已归一化
STUDENT_RANK - 学生在修读该课程时的大学排名（大一、大二等）
Age - 学生修读该课程时的年龄，已归一化
GENDER_DESC - 学生自我识别的性别
URM - 未充分代表的少数族裔状态
Student.22 - 学生指标
Student.01
Student.36
Student.04
Student.05
Student.23
Student.15
Student.28
Student.27
Student.10
Student.11
Student.12
Student.13
Student.14
Student.30
Student.16
Student.18
Student.31
Student.19
Student.20
Student.03
Student.09
Student.26
Student.21
Student.02
Student.06
Student.32
Student.08
Student.29
Student.07
Student.17
Student.25
Student.33
Student.34
Student.24
Student.35
SubjECON - 课程指标
SubjMGT
SubjFIN
SubjENGL
SubjACCT
SubjCIS
SubjMKT
SubjCOMM
SubjMATH
SubjBIOL
SubjBLAW
SubjHIST
SubjAE
SubjPSY
SubjMUS
SubjELCT
SubjPE
SubjART
SubjPOLS
SubjSPAN
Instructor.04 - 教师指标
Instructor.08
Instructor.07
Instructor.03
Instructor.16
Instructor.05
Instructor.12
Instructor.02
Instructor.06
Instructor.11
Instructor.13
Instructor.10
Instructor.01
Instructor.14
Instructor.09
Instructor.15
Advisor.02 - 导师指标
Advisor.01
Advisor.04
Advisor.11
Advisor.05
Advisor.09
Advisor.07
Advisor.03
Advisor.06
Advisor.10
Advisor.08
Advisor.12