TY - JOUR
T1 - Interval estimate for the X-intercept of a straight line
T2 - A nonlinear approach
AU - Schechtman, E.
AU - Spiegelman, C.
AU - Moran, N.
PY - 1998
Y1 - 1998
N2 - We show that a non-linear representation of a straight line model provides an opportunity for achieving reasonable interval estimates of the X-intercept when straight line model methods fail to allow such an interval. Further, we show that non-linear methods are reasonable competitors to the standard straight line method in Graybill (1976), even when the straight line method gives an estimate. The advantage of the non-linear approach is that the intervals may be directly obtained by using standard statistical packages such as SAS®. No extra programming is required.
AB - We show that a non-linear representation of a straight line model provides an opportunity for achieving reasonable interval estimates of the X-intercept when straight line model methods fail to allow such an interval. Further, we show that non-linear methods are reasonable competitors to the standard straight line method in Graybill (1976), even when the straight line method gives an estimate. The advantage of the non-linear approach is that the intervals may be directly obtained by using standard statistical packages such as SAS®. No extra programming is required.
KW - Asymptotics simulation
KW - Calibration
KW - Inverse regression
KW - Least squares
UR - http://www.scopus.com/inward/record.url?scp=0346967405&partnerID=8YFLogxK
U2 - 10.1080/03610919808813527
DO - 10.1080/03610919808813527
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:0346967405
SN - 0361-0918
VL - 27
SP - 1093
EP - 1115
JO - Communications in Statistics Part B: Simulation and Computation
JF - Communications in Statistics Part B: Simulation and Computation
IS - 4
ER -