Tools for IPAG Courses.

IPAG

The IPAG package provides a set of simple and pedagogical tools for statistical inference in R. It focuses on confidence intervals and concise linear regression summaries, with clear defaults and readable outputs.

The package is primarily intended for teaching purposes and applied work in economics, social sciences, and related fields.

Overview

IPAG provides wrapper functions around standard R statistical procedures, including:

• Student t tests for means and differences in means
• Exact binomial confidence intervals for proportions
• Confidence intervals for odds ratios from 2×2 contingency tables based on Fisher’s exact test
• Linear regression summaries based on ordinary least squares

All confidence intervals are computed at the 99% level by default, with the option to specify alternative confidence levels.

Installation

You can install the development version of IPAG from GitHub:

# install.packages("remotes")
remotes::install_github("gpiaser/IPAG")

Once available on CRAN, you will be able to install it with:

install.packages("IPAG")

Usage

Confidence intervals for a mean

mean_ci() computes a confidence interval for a population mean using a Student t test.

library(IPAG)
x <- c(4.2, 5.1, 6.3, 5.8, 4.9)
mean_ci(x)

Difference in means

mean_diff_ci() computes a confidence interval for the difference in means between two samples, either independent or paired.

x <- c(5.1, 4.9, 6.2, 5.8, 5.4)
y <- c(4.8, 4.7, 5.9, 5.2, 5.0)
mean_diff_ci(x, y)

Proportion

prop_ci() computes an exact confidence interval for a population proportion using binom.test()`.

prop_ci(trials = 100, successes = 45)

Odds ratio

oddsratio_ci() computes an exact confidence interval for the odds ratio from a 2x2 contingency table, based on Fisher's exact test.

oddsratio_ci(a = 12, b = 5, c = 4, d = 15)

Linear regression

linear_regress() provides a concise summary of a linear regression fitted with lm(). It reports the adjusted R-squared, the p-value of the overall Fisher test, and a table of estimated coefficients with confidence intervals and significance indicators.

data(Housing)
linear_regress(MEDV ~ RM + LSTAT, data = Housing)

Datasets

The package includes several datasets commonly used for illustration and teaching:

• Housing — hedonic housing prices (Harrison and Rubinfeld, 1978)
• Beauty — teaching evaluations and perceived attractiveness
• McKinsey — teacher compensation and student performance
• ContentMarketing — determinants of content marketing effectiveness
• Bosnia — microcredit program data from Bosnia and Herzegovina, used for impact evaluation
• covid19 — cross-country COVID-related indicators

Each dataset is documented and can be loaded using data().

Design principles

Transparency

All functions rely on well-established R functions such as t.test(), binom.test(), lm(), and fisher.test().

Consistency

Functions computing confidence intervals follow a consistent naming convention and return objects with simple S3 print methods.

Readability

Outputs are designed to be interpretable without requiring advanced knowledge of R object internals.

Intended use

IPAG is designed for instructional and applied contexts where clarity and interpretability are prioritized over extensibility or computational efficiency. It is not intended as a substitute for more comprehensive statistical packages.

License

This package is distributed under the MIT License.