. 2019 Apr 2;79(5):931–961. doi: 10.1177/0013164419839439

Table 3.

Four Different Type of Data Sets Used in Training Different XGBoost Models for a Hypothetical Three-Item Test.

	Input features (predictors)	Target output (1 = flagged, 0 = nonflagged)
Dichotomous response data	$(\begin{matrix} 1 & 0 & 1 \\ 1 & 1 & 1 \\ . & . & . \\ 0 & 0 & 1 \\ 0 & 1 & 1 \end{matrix})$	$(\begin{matrix} 1 \\ 0 \\ . \\ 0 \\ 0 \end{matrix})$
Nominal response data	$(\begin{matrix} A & A & B \\ A & B & B \\ . & . & . \\ B & C & B \\ D & B & B \end{matrix}) = (\begin{matrix} 1 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0 \\ 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 1 & 0 & 0 \\ . & . & . & . & . & . & . & . & . & . & . & . \\ 0 & 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 & 1 & 0 & 0 & 0 & 1 & 0 & 0 \end{matrix})$	$(\begin{matrix} 1 \\ 0 \\ . \\ 0 \\ 0 \end{matrix})$
Dichotomous response data + Response time data	$(\begin{matrix} 1 & 0 & 1 & 25 & 38 & 20 \\ 1 & 1 & 1 & 32 & 46 & 23 \\ . & . & . & . & . & . \\ 0 & 0 & 1 & 26 & 37 & 51 \\ 0 & 1 & 1 & 32 & 15 & 10 \end{matrix})$	$(\begin{matrix} 1 \\ 0 \\ . \\ 0 \\ 0 \end{matrix})$
Nominal response data + Response time data	$(\begin{matrix} 1 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 25 & 38 & 20 \\ 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 1 & 0 & 0 & 32 & 46 & 23 \\ . & . & . & . & . & . & . & . & . & . & . & . & . & . & . \\ 0 & 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 1 & 0 & 0 & 26 & 37 & 51 \\ 0 & 0 & 0 & 1 & 0 & 1 & 0 & 0 & 0 & 1 & 0 & 0 & 32 & 15 & 10 \end{matrix})$	$(\begin{matrix} 1 \\ 0 \\ . \\ 0 \\ 0 \end{matrix})$