This project analyzes the Students Performance in Exams dataset from Kaggle.

The dataset contains student exam scores in:

• Math
• Reading
• Writing

It also includes student background variables such as gender, lunch type, parental education, and test preparation course status.

The main questions are:

1. Do students who complete test preparation have higher average scores?
2. Is math score related to reading score?
3. Does lunch type appear related to average student performance?

The dataset comes from Kaggle:

Students Performance in Exams

The dataset includes information about student exam scores and background factors.

## # A tibble: 6 × 10
##   gender race_ethnicity parental_level_of_education lunch test_preparation_cou…¹
##   <chr>  <chr>          <chr>                       <chr> <chr>                 
## 1 female group B        bachelor's degree           stan… none                  
## 2 female group C        some college                stan… completed             
## 3 female group B        master's degree             stan… none                  
## 4 male   group A        associate's degree          free… none                  
## 5 male   group C        some college                stan… none                  
## 6 female group B        associate's degree          stan… none                  
## # ℹ abbreviated name: ¹​test_preparation_course
## # ℹ 5 more variables: math_score <dbl>, reading_score <dbl>,
## #   writing_score <dbl>, average_score <dbl>, passed_math <chr>

The main variables are:

• gender
• race_ethnicity
• parental_level_of_education
• lunch
• test_preparation_course
• math_score
• reading_score
• writing_score

I also created a new variable:

students <- students %>%
  mutate(
    average_score = (math_score + reading_score + writing_score) / 3
  )

The dataset was already mostly clean.

I renamed the columns to make them easier to use in R.

names(students) <- tolower(names(students))
names(students) <- gsub("[^a-z0-9]+", "_", names(students))
names(students) <- gsub("_$", "", names(students))

## # A tibble: 1 × 6
##   mean_math mean_reading mean_writing mean_average sd_average number_of_students
##       <dbl>        <dbl>        <dbl>        <dbl>      <dbl>              <int>
## 1      66.1         69.2         68.1         67.8       14.3               1000

## # A tibble: 2 × 6
##   test_preparation_course mean_math mean_reading mean_writing mean_average count
##   <chr>                       <dbl>        <dbl>        <dbl>        <dbl> <int>
## 1 completed                    69.7         73.9         74.4         72.7   358
## 2 none                         64.1         66.5         64.5         65.0   642

## [1] 0.8175797

There is a positive relationship between math score and reading score.

Students who score higher in math often also tend to score higher in reading.

I used a simple linear regression model to predict reading score from math score.

model <- lm(reading_score ~ math_score, data = students)
summary(model)

## 
## Call:
## lm(formula = reading_score ~ math_score, data = students)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -26.2905  -5.8011   0.1139   6.0341  21.4117 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 17.14181    1.19000   14.40   <2e-16 ***
## math_score   0.78723    0.01755   44.85   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 8.411 on 998 degrees of freedom
## Multiple R-squared:  0.6684, Adjusted R-squared:  0.6681 
## F-statistic:  2012 on 1 and 998 DF,  p-value: < 2.2e-16

## `geom_smooth()` using formula = 'y ~ x'

I used a two-sample t-test to compare average scores between students who completed test preparation and students who did not.

t_test_result <- t.test(
  average_score ~ test_preparation_course,
  data = students
)

t_test_result

## 
##  Welch Two Sample t-test
## 
## data:  average_score by test_preparation_course
## t = 8.5945, df = 791.84, p-value < 2.2e-16
## alternative hypothesis: true difference in means between group completed and group none is not equal to 0
## 95 percent confidence interval:
##  5.887734 9.373305
## sample estimates:
## mean in group completed      mean in group none 
##                72.66946                65.03894

The t-test compares the mean average score between the two groups:

• Students who completed test preparation
• Students who did not complete test preparation

If the p-value is small, it suggests that the average scores are different between the groups.

## # A tibble: 6 × 3
##   parental_level_of_education mean_average count
##   <chr>                              <dbl> <int>
## 1 master's degree                     73.6    59
## 2 bachelor's degree                   71.9   118
## 3 associate's degree                  69.6   222
## 4 some college                        68.5   226
## 5 some high school                    65.1   179
## 6 high school                         63.1   196

The main findings are:

• Students who completed test preparation tended to have higher average scores.
• Math and reading scores had a positive relationship.
• Lunch type appeared related to score differences.
• Parental education level also showed some differences in average score.

This project used data cleaning, exploratory data analysis, visualization, and simple statistical methods.

The analysis suggests that student performance is related to several factors, especially test preparation, lunch type, and performance across other subjects.

The results do not prove causation, but they show useful patterns in the data.