Table 1. Type I error control of different methods under null simulations.

Type I Error (alpha = 0.05, 0.005)	Genetic Covariance			Environmental Covariance
	One Study	Two Study	Overlapped	One Study	Overlapped
GECKO	2% (0.2%)	2% (0.8%)	2% (0.6%)	2% (0.2%)	3% (0.5%)
LDSC	3% (0.3%)	1% (0)	2% (0.1%)	NA	NA
GNOVA	37% (17.6%)	7% (0.8%)	27% (10.9%)	NA	NA

The table displays the type I error of different methods at a nominal p-value threshold of 0.05 and 0.005 (in parentheses) for testing the genetic covariance (first three columns) or the environmental covariance (last two columns) under null simulations with three different study designs. The three different study designs include the one study design, two separate study design, and two study design with partially overlapped individuals. LDSC and GNOVA cannot be used for testing the environmental covariance so NAs are shown in the corresponding cells.