Activity

Participants move from the supervised-evaluation motivation into hands-on 2PL ICCs, manipulating difficulty and discrimination values directly.

Browser Lab

This page now includes executable Python cells directly in the site. Students can adjust 2PL item parameters, rerun cells, add scratch cells, and compare curves without leaving the workshop page.

In this section, the 2PL ability parameter theta is real-valued. The interactive lab plots the ICC over theta in [-3, 3], which is a common practical window for visualizing the central part of a latent ability scale. Item difficulty is also real-valued and acts as the location of the curve on that axis. Discrimination is a slope parameter; the standard case is positive discrimination, although the input fields allow custom numeric values for exploration.

Shared Helpers You Can Use

The browser notebook, the local notebook, and Colab can all import the same workshop helpers directly. You do not need to rewrite these utilities during the activity.

from utils.handson import (
    binary_irt_probability,
    make_binary_item_bank,
    plot_binary_iccs,
)

• binary_irt_probability computes the 2PL probability curve for chosen ability, difficulty, and discrimination values.
• make_binary_item_bank creates a small starter item bank that is ready to plot and edit.
• plot_binary_iccs plots several ICCs at once so you can compare items quickly in the activity notebook or locally.

Inside the browser notebook editor, hovering one of these helper names now shows its docstring.

Notebook

• Activity notebook: 02_01_activities.ipynb
• Goal: connect 2PL item parameters to the geometry of the curves

What Participants Should Do

• change item difficulty;
• change item discrimination;
• compare 2PL ICC shapes and discuss what they mean.