Course (Part 575)

Lesson

Topic: Welcome to GREerbal Precision — R fundamentals, visualization, and basic modeling

Learning Outcomes

• Install and set up R and RStudio; create and inspect data frames.
• Compute descriptive statistics (mean, SD) and produce ggplot2 visualizations.
• Fit and interpret a simple linear regression with lm().
• Recognize modeling assumptions and common pitfalls (overfitting, missing data).
• Communicate analytic findings precisely and concisely.

Conceptual Explanation

GREerbal Precision blends verbal reasoning with practical R skills. Start by creating a tidy data frame and using summary statistics to describe your dataset. Visualize categories and numeric relationships in ggplot2. Fit simple models to explore predictive relationships but always check assumptions.

Core R pattern (minimal example):

# Create a simple data frame
data <- data.frame(
  Name = c("Alice", "Bob", "Charlie", "Diana"),
  Age  = c(25, 30, 35, 40),
  Score= c(88, 92, 95, 90)
)
print(data)
# Basic ggplot2 barplot
library(ggplot2)
ggplot(data, aes(x = Name, y = Score, fill = Name)) +
  geom_bar(stat = "identity") + theme_minimal()
# Descriptive stats
mean_score <- mean(data$Score)
sd_score   <- sd(data$Score)
cat("Mean:", mean_score, "SD:", sd_score, "\n")
# Linear regression
model <- lm(Score ~ Age, data = data)
summary(model)

Interpretation tip: coefficients indicate expected change per unit; R-squared shows explained variance but does not prove causation.

Bilingual (EN → বাংলা)

EN: Create a data frame in R to store variables such as name, age, and score.

বাংলা: R-এ একটি ডেটা ফ্রেম তৈরি করে নাম, বয়স এবং স্কোরের মতো ভেরিয়েবল সংরক্ষণ করুন।

EN: Use ggplot2 to visualize results and make patterns obvious.

বাংলা: ফলাফল দৃশ্যায়নের জন্য ggplot2 ব্যবহার করুন যাতে প্যাটার্ন সহজে দেখা যায়।

EN: Always check model assumptions before trusting predictions.

বাংলা: ভবিষ্যদ্বাণী বিশ্বাস করার আগে মডেলের ধরা-ধাক্কা গুলো পরীক্ষা করুন।

Important Notes

• Always inspect raw data (str(), head(), summary()) before analysis.
• Document every transformation and why it was applied.
• Impute or filter missing values with justification—don’t hide exclusions.
• Use reproducible code chunks and set seed for simulations (set.seed()).
• Visualize residuals and check homoscedasticity and normality for lm().

Exercises → True / False / Fill in the Blanks / MCQ / Matching / Descriptive / Creative

Activities

1. Run the starter script to create the data frame and produce the bar plot.
2. Compute mean and SD for Score; interpret.
3. Fit lm(Score ~ Age) and write a one-sentence interpretation of the Age coefficient.

Vocabulary

• Data frame — tabular R object holding observations.
• Homoscedasticity — constant error variance assumption.
• Residual — observed minus predicted value.
• R-squared — proportion variance explained by model.

Homework

1. Install tidyverse, run the starter script, and save a PNG of the plot.
2. Collect a small dataset (n ≥ 8) and compute mean, median, SD; submit summary table.

Practice Task

Using the iris dataset, compute mean Sepal.Length by Species and visualize with a bar chart.

library(dplyr); library(ggplot2)
iris %>% group_by(Species) %>% summarise(Mean_Sepal = mean(Sepal.Length)) -> s
ggplot(s, aes(x=Species, y=Mean_Sepal, fill=Species)) + geom_col() + theme_minimal()

Exercise Types: Questions & Answers

True / False (3)

1. lm() in R automatically proves causation when p < 0.05. — True / False
   Answer

   False. A significant p-value shows association given model assumptions, not causation; consider confounding and design.

2. A bar plot with geom_bar(stat="identity") requires a y aesthetic. — True / False
   Answer

   True. When using stat="identity", you supply the y values; otherwise the count is computed.

3. Mean and median are equally sensitive to outliers. — True / False
   Answer

   False. The mean is more sensitive to outliers; the median is robust.

Fill in the Blanks (3)

1. To view the first rows of a data frame use ___.
Answer

head()

2. To compute standard deviation in R use ___.
Answer

sd()

3. In ggplot2, map aesthetics inside ___ (function name) to connect variables to visual properties.
Answer

aes()

Multiple Choice (3)

1. Which function fits a linear model in base R?
   A) lm()   B) glmnet()   C) kmeans()   D) plot()
   Answer

   A) lm()

2. Which ggplot2 geom is appropriate for scatter plots?
   A) geom_bar()   B) geom_point()   C) geom_boxplot()   D) geom_density()
   Answer

   B) geom_point()

3. Which R package is part of the tidyverse for data manipulation?
   A) dplyr   B) lattice   C) car   D) randomForest
   Answer

   A) dplyr

Matching (3 pairs)

Match R term → Description

Descriptive (short)

Explain in 2–3 sentences why plotting residuals is important after fitting a linear model.

Creative

Design a one-paragraph outreach blurb (≤ 50 words) describing why GREerbal Precision helps students merge verbal reasoning with data skills.

Further Resources

• R for Data Science — Hadley Wickham & Garrett Grolemund
• Install packages: install.packages("tidyverse")
• Community: RStudio Community, Stack Overflow (tag: r)