Exploring Academic Diversity: SAS Analysis of Midterm Grades by Race and Gender

In our analysis, we delve into the influence of race and gender on students' academic performance, employing the powerful SAS software for a robust examination. We dissect race-based grade distributions, test gender-based proportions, assess normality in reading scores, and scrutinize subject scores by gender. This comprehensive investigation seeks to shed light on whether these factors have a significant impact on midterm grades, providing valuable insights into educational equity and helping institutions make informed decisions to support their students.

Problem Description

In a comprehensive statistical analysis assignment using SAS of midterm grades, we explore the influence of race and gender on students' performance across various subjects. We aim to answer whether race or gender significantly impacts midterm grades. Using SAS software, we analyze race-based grade distributions, test for gender-based proportions, assess normality in reading scores, and conduct t-tests to compare subject scores by gender. Our goal is to understand the factors that affect academic performance and draw meaningful conclusions.

Part 1: Analysis by Race

Statistics for Midterm Grades Based on Race

Table 1:Race-Based Midterm Grade Distribution

Chi-Square Test for Specified Proportions

• Chi-Square: 0.0519
• Degrees of Freedom (DF): 2
• p-value (Pr>ChiSq): 0.9744

The p-value (0.9744) is greater than the significance level (5%), indicating that there is no significant difference between the race percentages and the specified values (0.45, 0.2, and 0.35, respectively).

Part 2: Analysis by Gender

Statistics for Midterm Grades Based on Gender

Table 2: Gender-Based Midterm Grade Distribution

Chi-Square Test for Equal Proportions

• Chi-Square: 1.4727
• Degrees of Freedom (DF): 1
• p-value (Pr>ChiSq): 0.2249

The p-value (0.2249) is greater than the significance level (5%), indicating that there is no significant difference between the proportions of males and females.

Part 3: Normal Distribution Test

Fitted Normal Distribution for Reading Scores

Table 3:Fitted Normal Distribution for Reading Scores

Goodness-of-Fit Tests for Normal Distribution

• Chi-squared Test
• Degrees of Freedom (DF): 4
• Statistic: 3.4895
• p-value (Pr> χ^2): 0.4795

The p-value (0.4795) is greater than the significance level (5%), suggesting that student reading scores follow a normal distribution with a mean of 47.8 and a standard deviation of 8.8.

Part 4: Independent Sample t-Test

Comparison of Scores by Subject and Gender

Table 4:Comparison of Scores by Subject and Gender (Pooled and Satterthwaite Variance)

Null Hypotheses:

• There is no difference between the mean scores of (read, write, math, science, socst) for males and females.

Alternative Hypotheses:

• There is a difference between the mean scores of (read, write, math, science, socst) for males and females.

For each subject (read, write, math, science, socst), two types of t-tests were conducted: Pooled (Equal Variance) and Satterthwaite (Unequal Variance).

The p-value for all t-tests conducted is greater than the significance level (0.05), indicating that there is no significant difference between the mean scores across genders.

Part 5: Test for Equality of Means

Comparison of Sciences and Mathematics Scores of Females

Table 5: Comparison of Sciences and Mathematics Scores of Females (Pooled and Satterthwaite Variance)

Sciences and Mathematics Scores of Females

• Pooled (Equal Variance)
• Degrees of Freedom (DF): 44
• t Value: 0.97
• p-value (Pr> |t|): 0.3390
• Satterthwaite (Unequal Variance)
• Degrees of Freedom (DF): 42.021
• t Value: 0.97
• p-value (Pr> |t|): 0.3393

The p-values (0.3390 and 0.3393) are both greater than the significance level (0.05), indicating that there is no significant difference between the population mean of mathematics and science scores among females.

Writing and Social Sciences Scores of Females

• Pooled (Equal Variance)
• Degrees of Freedom (DF): 44
• t Value: 0.11
• p-value (Pr> |t|): 0.9129
• Satterthwaite (Unequal Variance)
• Degrees of Freedom (DF): 43.84
• t Value: 0.11
• p-value (Pr> |t|): 0.9129

The p-values (0.9129 for both tests) are greater than the significance level (0.05), indicating no significant difference between the population mean of writing and social science scores among females.

In conclusion, the analysis shows that there are no significant differences in midterm grades based on race or gender. Additionally, the scores follow a normal distribution, and there are no significant differences in mean scores across subjects and genders.