Multiple R-squared for Women’s college GPA based on HSGPA and SAT Score:
## [1] 0.3223841
Multiple R-squared for Men’s college GPA based on HSGPA and SAT Score:
## [1] 0.1570066
Conclusions from Data:
College GPA and HSGPA appear to have a weak effect on the First Year
College GPA. Despite this, R-squared values for the linear models of
both genders suggests that women’s success in college is more impacted
by their HSGPA and SAT scores than men’s.
Code:
library(mosaic)
library(dplyr)
library(ggplot2)
FirstYearGPA <- read.csv("C:/Users/casey/Downloads/FirstYearGPA.csv")
df1 <- FirstYearGPA %>%
mutate(
Gender = ifelse(Male == "1", "Male",
ifelse(Male == "0", "Female", "NA")),
SAT = SATV + SATM,
SatCat = ifelse(SAT >= 1239, "Top 50%",
ifelse(SAT <=1239, "Bottom 50%", "NA"))
)
female <- filter(df1, Gender == "Female")
male <- filter(df1, Gender == "Male")
ggplot(data = df1, aes(x = SAT, y = GPA)) +
geom_point() +
aes(colour = HSGPA) +
facet_wrap(~Gender, ncol = 4) +
stat_smooth(method = lm, se = FALSE, color = "black") +
theme_bw() +
theme(panel.grid.minor = element_blank(), panel.grid.major = element_blank()) +
labs(title = "SAT Score and HS GPA as a Potential Predictor of College GPA", x = "SAT Score", y = "College GPA")
lmfemale <- lm(GPA ~ SAT + HSGPA, data = female)
summary(lmfemale)$r.squared
lmmale <- lm(GPA ~ SAT + HSGPA, data = male)
summary(lmmale)$r.squared