cjdiag: Diagnostic Tools for Conjoint Survey Experiments

Tools for attribute-level importance and attendance in conjoint survey experiments — which attribute levels drive choices, how they rank, and which ones respondents ignore.

Entry Points

• cj_fit() — Fits one of 5 methods: "forest", "tree", "crt", "nmm", "marginal_r2"

Common Workflow

# 1. Fit a model
rf <- cj_fit(outcome ~ attr1 + attr2, data = df, method = "forest")

# 2. View results
print(rf)
importance(rf)

# 3. Visualize
plot(rf)
plot(rf, palette = "colorblind", group_by_attribute = TRUE)

Customization

All plot methods support base_size, colors, palette, theme, label_wrap, attribute.names, and level.names parameters. Use set_cjdiag_theme() and set_cjdiag_labels() to set global defaults.

See also

Useful links:

• https://dkarpa.github.io/cjdiag/

• https://github.com/dkarpa/cjdiag

• Report bugs at https://github.com/dkarpa/cjdiag/issues

Author

Maintainer: David Karpa davidfkarpa@gmail.com (ORCID) [copyright holder]