Secondary Analysis under Cohort Sampling Designs Using Conditional Likelihood

Table 5

Maximum likelihood estimates (standard errors) of the parameters 𝛽 of the BMI model 𝑌_𝑖∣𝑏_𝑖,𝑍_𝑖,𝛽∼NPQM(𝛽₀+𝛽₁𝑏_𝑖+𝛽₂1_{𝑍𝑖=0},𝛽₃,𝛽₄,𝛽₅).

Likelihood	𝑘	𝛽̂0	𝛽̂1	𝛽̂2	𝛽̂3	log(𝛽̂3/√𝜋)	𝛽̂4	𝛽̂5	log𝛽̂5
Conditional	100	26.91 (0.14)	−0.07 (0.02)	−0.64 (0.27)	4.27	0.88 (0.02)	0.14 (0.03)	0.37	−1.00 (0.05)
(6.2)	1000	26.92 (0.14)	−0.07 (0.02)	−0.64 (0.27)	4.27	0.88 (0.02)	0.14 (0.03)	0.37	−1.00 (0.05)
	5000	26.92 (0.14)	−0.07 (0.02)	−0.64 (0.27)	4.27	0.88 (0.02)	0.14 (0.03)	0.37	−1.00 (0.05)
Conditional	100	26.93 (0.13)	−0.07 (0.02)	−0.63 (0.26)	4.25	0.87 (0.02)	0.14 (0.03)	0.37	−1.00 (0.05)
(6.3)	1000	26.93 (0.13)	−0.07 (0.02)	−0.63 (0.26)	4.25	0.87 (0.02)	0.14 (0.03)	0.37	−1.00 (0.05)
	5000	26.93 (0.13)	−0.07 (0.02)	−0.63 (0.26)	4.25	0.87 (0.02)	0.14 (0.03)	0.37	−1.00 (0.05)
Full	—	26.94 (0.07)	−0.07 (0.01)	−0.60 (0.25)	4.06	0.83 (0.01)	0.13 (0.01)	0.35	−1.06 (0.03)
Weighted	—	26.94 (0.15)	−0.06 (0.03)	−0.68 (0.33)	4.30	0.88 (0.04)	0.17 (0.05)	0.41	−0.89 (0.11)