Improving statistical inference on pathogen densities estimated by quantitative molecular methods: malaria gametocytaemia as a case study

Table 1 Summary of the homoscedastic linear model calibration curves fitted by ordinary least squares

Curve	Intercept ^a , β ₀ (SE ^b )	Slope, β ₁ (SE)	Variance, σ ²	R ²	g ^c
j = 1	30.87 (0.96)	−4.42 (0.56)	5.58	0.94	0.13
j = 2	32.46 (1.90)	−4.79 (1.11)	21.75	0.82	0.42
j = 3	30.88 (0.64)	−3.80 (0.38)	2.47	0.96	0.08
j = 4	30.82 (2.35)	−4.59 (1.37)	33.00	0.74	0.69
j = 5	27.00 (0.51)	−2.48 (0.30)	1.54	0.95	0.11
j = 6	33.89 (1.16)	−2.69 (0.68)	8.14	0.79	0.50
j = 7	31.21 (0.53)	−3.49 (0.31)	1.66	0.97	0.06
j = 8	24.09 (0.46)	−2.40 (0.27)	1.25	0.95	0.09
j = 9	24.28 (0.91)	−2.98 (0.53)	4.93	0.89	0.24
j = 10	30.25 (1.62)	−4.73 (0.95)	15.78	0.86	0.31
j = 11	27.71 (0.58)	−3.15 (0.34)	2.00	0.96	0.09
j = 12	20.85 (0.58)	−2.51 (0.34)	2.00	0.93	0.14

^aCalibrators, x _ij, were centered about their mean, \( \overline{x}, \) ensuring that ‘intercept’ terms correspond to the respective estimates at \( {x}_{ij}=\overline{x} \).
^bStandard error.
^cCalculated using a Student’s critical t value at a significance level of 5% and n + m −3 = 4 + 1–3 degrees of freedom (8).

ISSN: 1471-2105