In a method comparison for cholesterol, linear regression of the new method (y) against the current method (x) gives y = 1.01x + 12 mg/dL with a correlation coefficient (r) of 0.998. What does this most strongly indicate about the new method?
1A constant systematic error of about +12 mg/dL✓ 정답
2Excess random error, because the intercept is not zero
3A proportional systematic error that increases with concentration
4No bias, because r is very close to 1.0
해설
A slope near 1.00 with an intercept of +12 mg/dL means the new method reads about 12 mg/dL higher across the range, a constant systematic error often caused by an interferent or blanking problem. A proportional error would appear as a slope clearly different from 1, and a high r reflects only how closely the data fit a line, not agreement between methods.