Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Fitting straight lines to experimental data

The American Journal of Physiology

|September 1, 1977

Summary

This summary is machine-generated.

Standard regression underestimates slope when both variables have errors. This study introduces a simple method using standard errors to find the best linear relationship (Y = AX + B), providing a more accurate slope and intercept.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Ovine amniotic fluid volume response to intra-amniotic balloon filling.

Placenta·2008

Same author

Expression of erythropoietin mRNA, protein and receptor in ovine fetal membranes.

Placenta·2001

Same author

Erythropoietin responses to progressive blood loss over 10 days in the ovine fetus.

American journal of physiology. Regulatory, integrative and comparative physiology·2001

Same author

Antibody-induced anemia in fetal sheep: model for hemolytic disease of the fetus and newborn.

Journal of the Society for Gynecologic Investigation·2001

Same author

Erythrocyte and erythropoietin responses to hemorrhage in the immature and near term ovine fetus.

American journal of obstetrics and gynecology·2001

Same author

Effect of laboratory acclimation on food and water consumption of pregnant sheep after fetal catheterization.

Contemporary topics in laboratory animal science·2001

Same journal

Blood coagulation in fish.

The American journal of physiology·2011

Same journal

Renal tubular reabsorption, metabolic utilization and isomeric fractionation of lactic acid in the dog.

The American journal of physiology·2010

Same journal

The inactivation of placental toxin by human serum.

The American journal of physiology·2010

Same journal

Adrenal function following ovariectomy in the rat.

The American journal of physiology·2010

Same journal

Capillary permeability; perfusion of frog and guinea pig hind limbs.

The American journal of physiology·2010

Same journal

Evaluation of protective measures against sunburn.

The American journal of physiology·2010

See all related articles

Area of Science:

Statistical modeling
Experimental research methodology

Background:

Standard least-squares regression analysis is commonly used to determine linear relationships (Y = AX + B).
Experimental data often involves errors in both measured variables, a condition not optimally handled by standard methods.

Purpose of the Study:

To address the underestimation of the slope (A) in linear regression when both variables contain errors.
To introduce a simple and accurate method for estimating the best linear relationship under these conditions.

Main Methods:

Examined the limitations of standard regression when both variables have unknown errors.
Proposed a new method calculating the slope as the ratio of standard errors (Y/X) and the intercept from mean values (B = Y - AX).
Demonstrated the slope estimate is also the conventional slope divided by the absolute correlation coefficient.

Related Experiment Videos

Main Results:

Standard least-squares regression underestimates the true slope (A) when errors exist in both variables.
The proposed method provides a slope estimate equal to the ratio of standard errors of Y and X.
The intercept is accurately determined using the mean values of X and Y.
The derived line represents a line of symmetry through the data points.

Conclusions:

A simple method exists to accurately determine the best linear relationship when both variables have errors.
This method offers a more robust alternative to standard regression in such prevalent experimental scenarios.
The resulting linear model provides a symmetrical representation of the data.