Student Performance: Map academic influences through Chi-Square association tests across subjects. Python, Scipy, Seaborn, Pandas.
Assignment
The purpose of this project is to delve into student data to uncover insights that could help understand and predict student success across different academic periods. As a contributor, you will explore the relationship between student grades and a myriad of demographic, social, and school-related factors. Your analysis will directly impact our ability to identify and address key influences on student performance.
Your Tasks:
-
Create Data Visualizations:
- Generate histograms to observe the distribution of grades (G1, G2, G3) and other numerical factors like age and study time.
- Construct box plots to spot outliers and understand the spread of the data.
- Use bar charts to compare the average grades across different categories such as gender, parental education level, and internet access.
- Your visualizations should help highlight trends and patterns that may influence student performance.
-
Perform Statistical Tests:
- Conduct Chi-Square Tests to investigate the association between categorical variables (e.g., gender, internet access) and student grades. For example, is there a statistical difference in grades between students with different levels of parental education?
- Apply t-tests to compare the mean grades between two different groups, such as students from urban versus rural areas.
- Use regression analysis to predict final grades based on various factors like study time and past failures.
- Document your findings, interpret the p-values, and discuss the statistical significance of your results.
Objectives:
- Correlation Analysis: Determine which factors are most strongly correlated with student grades.
- Predictive Modeling: Build a model that can predict a student's final grade based on their background and school-related activities.
- Insight Generation: Provide actionable insights for schools to help improve student outcomes based on your findings.
Tools You Might Need:
- For Visualization: Matplotlib, Seaborn, or any other Python library that you are comfortable with.
- For Statistical Analysis: SciPy for conducting statistical tests, and statsmodels or Scikit-learn for any predictive modeling.
Data Description
The dataset consists of two files, student-mat.csv and student-por.csv, which include various features that describe student demographics, family background, school-related factors, and academic performance. Below is a detailed breakdown of each variable:
Student Information
- school (Categorical): The school attended by the student ('GP' - Gabriel Pereira or 'MS' - Mousinho da Silveira).
- sex (Binary): Student's gender ('F' - Female, 'M' - Male).
- age (Integer): Student's age (ranging from 15 to 22 years old).
- address (Categorical): Student's home address type ('U' - Urban, 'R' - Rural).
Family Background
- famsize (Categorical): Family size ('LE3' - Less or equal to 3, 'GT3' - Greater than 3).
- Pstatus (Categorical): Parent's cohabitation status ('T' - Living together, 'A' - Apart).
- Medu (Integer): Mother's education level (0 - None, 1 - Primary, 2 - 5th to 9th grade, 3 - Secondary, 4 - Higher education).
- Fedu (Integer): Father's education level (0 - None, 1 - Primary, 2 - 5th to 9th grade, 3 - Secondary, 4 - Higher education).
- Mjob (Categorical): Mother's job ('teacher', 'health', 'services', 'at_home', 'other').
- Fjob (Categorical): Father's job ('teacher', 'health', 'services', 'at_home', 'other').
- guardian (Categorical): Student’s guardian ('mother', 'father', 'other').
Academic and School-Related Factors
- reason (Categorical): Reason for choosing the school ('home', 'reputation', 'course', 'other').
- traveltime (Integer): Home to school travel time (1 - <15 min, 2 - 15-30 min, 3 - 30 min-1 hour, 4 - >1 hour).
- studytime (Integer): Weekly study time (1 - <2 hours, 2 - 2-5 hours, 3 - 5-10 hours, 4 - >10 hours).
- failures (Integer): Number of past class failures (numeric: n if 1<=n<3, else 4).
Educational Support & Extracurricular Activities
- schoolsup (Binary): Extra educational support ('yes' or 'no').
- famsup (Binary): Family educational support ('yes' or 'no').
- paid (Binary): Extra paid classes for Math/Portuguese ('yes' or 'no').
- activities (Binary): Participation in extracurricular activities ('yes' or 'no').
- nursery (Binary): Attended nursery school ('yes' or 'no').
- higher (Binary): Aspiration to take higher education ('yes' or 'no').
- internet (Binary): Internet access at home ('yes' or 'no').
- romantic (Binary): In a romantic relationship ('yes' or 'no').
Social and Lifestyle Factors
- famrel (Integer): Quality of family relationships (1 - Very bad to 5 - Excellent).
- freetime (Integer): Free time after school (1 - Very low to 5 - Very high).
- goout (Integer): Going out with friends (1 - Very low to 5 - Very high).
- Dalc (Integer): Workday alcohol consumption (1 - Very low to 5 - Very high).
- Walc (Integer): Weekend alcohol consumption (1 - Very low to 5 - Very high).
- health (Integer): Current health status (1 - Very bad to 5 - Very good).
- absences (Integer): Number of school absences (0 to 93).
Academic Performance (Target Variables)
- G1 (Integer): First period grade (0 to 20).
- G2 (Integer): Second period grade (0 to 20).
- G3 (Integer): Final grade (0 to 20, target variable).
Practicalities
- The data covers a wide range of factors, from personal demographics to academic records.
- Analyze the datasets to understand how these factors correlate with student performance in Mathematics and Portuguese language.