Convert Confidence Level to Variance Inflation Factor

Maps qualitative confidence levels to quantitative VIF values. The VIF determines prior variance as: Var(K) = VIF * (mu_K - 1).

Usage

confidence_to_vif_fit(confidence)

Arguments

confidence

Character; one of "low", "medium", or "high".

Value

Numeric; the corresponding VIF value.

Details

The mapping is:

• "low": VIF = 5.0 (very high uncertainty about cluster count)

• "medium": VIF = 2.5 (moderate uncertainty)

• "high": VIF = 1.5 (strong prior belief)

These values are chosen to provide a wide range of uncertainty levels. VIF = 5 corresponds to "I have little idea how many clusters there are."

See also

vif_to_variance_fit, DPprior_fit

Examples

if (FALSE) { # \dontrun{
confidence_to_vif_fit("low")     # 5.0
confidence_to_vif_fit("medium")  # 2.5
confidence_to_vif_fit("high")    # 1.5
} # }