Chapter 1 Introduction

This manual explains how to build an autograder for UCSB ECON 145 coding assignments.

WHAT IS AN AUTOGRADER?

In this manual, an autograder, or autograder program, is an R script that grades a student’s R-based assignment and provides question-specific feedback. It runs alongside a set of supporting R scripts, collectively known as the autograder infrastructure.

See Autograder Conventions for more details.

In ECON 145 assignments, coding questions are classified as either Public or Private.

• For any Public Question, we write custom “Checks” – lines of code that test for specific answer attributes – that evaluate the student’s answer and return constructive and dynamic feedback.

• For any Private Question, we use a set of premade functions that evaluate the student’s answer and return necessary and basic feedback.

Accordingly, most of this manual focuses on writing Checks for Public Questions.

We will look at grading four types of data structures: tibbles, vectors, lists, and plots. Moreover, as the majority of ECON 145 assignment questions are tibble-based, we direct our attention to grading tibbles.

For fellow ECON 145 TAs, we recommend skimming through 00-manual1.pdf and how_to_create_an_autograder_2024.pdf for an overview of the autograder infrastructure.

Note:

Unless noted, all references and examples come from the Fall 2024 ECON 145 course.