rho_c = 2 s_xy / (s_x^2 + s_y^2 + (xbar - ybar)^2) with moments on divisor n
(Lin 1989; Hedderich, Sachs & Reynarowych 2023, Sec. 6.16.4), and Lin's
interval on z = atanh(rho_c) with se = sqrt(((1 - r^2) rho_c^2 (1 - rho_c^2) /
r^2 + 2 rho_c^3 (1 - rho_c) u^2 / r - rho_c^4 u^4 / (2 r^2)) / (n - 2)) / (1 -
rho_c^2), u = (ybar - xbar) / (s_x^2 s_y^2)^(1/4): DescTools' CCC(ci =
"z-transform"). At |rho_c| = 1 the interval is the point itself.
Usage
morie_concordance_corr(x, y, confidence = 0.95)
Arguments
- x, y
Paired numeric measurements.
- confidence
Interval level.
Value
Named list: estimate, ci_lower, ci_upper, se, n, pearson_r,
asymptotic_ci, scale_shift, location_shift, C_b.
References
Lin, L. I.-K. (1989). Biometrics 45, 255-268; correction (2000)
Biometrics 56, 324-325.
Examples
morie_concordance_corr(c(1, 2, 3.2, 4.1, 5), c(1.1, 2.3, 2.9, 4.4, 5.2))$estimate
#> [1] 0.9847995
