Table 4.

Results of simulation study for the SHP2 inhibition experiment with 10 mice per group and 7 measurements per mouse³⁰.

Scenario	ρ	True ${\beta }_2$	Model	Mean estimated ${\beta }_2$ (IQR)	Coverage	Power	Type I error**
1	0	−0.008	Ind	−0.0080 (−0.0118, −0.0041)	0.9477	0.3020	0.0520
			AR-1	−0.0080 (−0.0117, −0.0041)	0.9503	0.2907	0.0483
			CS	−0.0080 (−0.0117, −0.0041)	0.9320	0.3180	0.0590
			IND-I	−0.0082 (−0.0145, −0.0020)	0.9540	0.1347	0.0543
			Mixed AR-1*	−0.0079 (−0.0115, −0.0042)	0.9659	0.2638	0.0385
2	0	−0.015	Ind	−0.0150 (−0.0187, −0.0114)	0.9510	0.7690
			AR-1	−0.0150 (−0.0187, −0.0114)	0.9550	0.7593
			CS	−0.0150 (−0.0187, −0.0113)	0.9410	0.7783
			IND-I	−0.0151 (−0.0213, −0.0090)	0.9570	0.3620
			Mixed AR-1*	−0.0148 (−0.0186, −0.0110)	0.9605	0.7125
3	0.5	−0.008	Ind	−0.0079 (−0.0123, −0.0034)	0.8967	0.3323	0.1067
			AR-1	−0.0079 (−0.0123, −0.0035)	0.9490	0.2187	0.0537
			CS	−0.0078 (−0.0125, −0.0033)	0.9090	0.2780	0.0907
			IND-I	−0.0077 (−0.0151, −0.0000)	0.8717	0.2063	0.1150
			Mixed AR-1*	−0.0076 (−0.0118, −0.0033)	0.9564	0.1891	0.0609
4	0.5	−0.015	Ind	−0.0149 (−0.0195, −0.0104)	0.8947	0.7220
			AR-1	−0.0150 (−0.0196, −0.0103)	0.9500	0.6040
			CS	−0.0150 (−0.0198, −0.0101)	0.9143	0.6510
			IND-I	−0.0150 (−0.0225, −0.0075)	0.8853	0.4163
			Mixed AR-1*	−0.0147 (−0.0194, −0.0101)	0.9511	0.5772
5§	0.96	−0.008	Ind	−0.0083 (−0.0157, −0.0010)	0.6522	0.4582	0.3427
			AR-1	−0.0082 (−0.0128, −0.0037)	0.9507	0.2534	0.0523
			CS	−0.0082 (−0.0134, −0.0032)	0.6776	0.5605	0.3047
			IND-I	−0.0082 (−0.0136, −0.0031)	0.6742	0.5462	0.3093
			Mixed AR-1*	−0.0076 (−0.0119, −0.0032)	0.9280	0.2139	0.0597
6	0.96	−0.015	Ind	−0.0148 (−0.0222, −0.0073)	0.6547	0.6523
			AR-1	−0.0148 (−0.0192, −0.0103)	0.9413	0.6230
			CS	−0.0148 (−0.0198, −0.0099)	0.6847	0.8423
			IND-I	−0.0148 (−0.0200, −0.0097)	0.6787	0.8220
			Mixed AR-1*	−0.0144 (−0.0190, −0.1000)	0.9272	0.5857

Covariance matrix structures include independence (Ind), autoregressive (AR-1) & compound symmetry (CS). IND-I corresponds to the model with independence covariance structure and a mouse indicator (fixed-effects model). Mixed AR-1 corresponds to the mixed-effects model with random intercept.

*The percentage of datasets for which the model did not converge was 20.7, 24.7, 32.6, 37, 48, 50 for Scenario 1, 2, 3, 4, 5, 6, respectively. For the scenarios for type I error evaluation, the associated percentages were 18.5, 32 and 44 for ρ of 0, 0.5 and 0.96, respectively.

**Type I error is derived from corresponding scenarios with ${\beta }_2$ = 0.

^§Scenarios in bold face reflect parameter values actually observed in the experiment.