Astronomy and Astrophysics – Astronomy
Scientific paper
Nov 1983
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=1983esasp.201..225s&link_type=abstract
In ESA Statist. Methods in Astron. p 225-229 (SEE N84-19196 09-89)
Astronomy and Astrophysics
Astronomy
Binary Stars, Data Correlation, Polarimetry, Stellar Models, Stellar Spectrophotometry, Confidence Limits, Parameter Identification, Signal To Noise Ratios
Scientific paper
Applying formal error techniques or results from linear model theory to nonlinear problems is discussed in relation to polarimetric modelling of binary stars. The least squares estimator for the inclination of the binary is shown to be severely biased towards high values when the signal to noise ratio is low, particularly for low inclinations. This biasing is obscured by selection effects that favor the detection of high inclination systems. Most close binary systems are known as such because they manifest spectrophotometric variations, but these are small when the binary inclination is low. Hence it is extremely probable that there is agreement between polarimetric and spectrophotometric determinations of binary inclinations, although this reflects no more than the noise level of polarimetric data. Methods of dealing with the bias in this estimation are discussed, stressing the importance of establishing confidence regions.
No associations
LandOfFree
Bias in polarimetric estimators does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Bias in polarimetric estimators, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Bias in polarimetric estimators will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1485337