Exponential models
In a nutshell
Experimental data often forms polynomial and exponential models in the form y=axn and y=abx. It is not practical to examine these non-linear trends. Logarithms can be used to code regression lines which practically define trends and allow for a more precise form of extrapolation.
The case where y=axn
Given a non-linear relationship of the form y=axn, you can apply the logarithm with base 10 (or any base) and use logarithm laws to obtain a linear relationship between log(y) and log(x):
ylog(y)=axn=log(axn)
log(y)=nlog(x)+log(a)
This equation is a linear relationship in the form Y=mX+c, where Y=log(y), X=log(x), m=n and c=log(a).
Example 1
The kinetic energy of a ball, denoted as K, is given by the formula K=21mv2 where m is the mass (kg), and v is the velocity (ms−1). An experiment is undertaken to measure the kinetic energy of the ball as its velocity increases with the results given in the table below to 1 decimal place. Using this data, find the mass of the ball to 2 decimal places. Do this by coding a regression line in the form Y=mX+c.
K(J) | 39.1 | 44.4 | 50.4 | 92.3 | 123.4 |
v (ms−1) | 22.9 | 24.4 | 26.0 | 35.2 | 40.7 |
The relationship between kinetic energy and velocity is in the form:
Ky=(21m)v2=(a)xn
Code this relationship to fit a regression line:
log(y)log(K)Y=nlog(x)+log(a)=2log(v)+log(21m)=mX+c
Create a table of log(K) against log(v) to 2 d.p.:
log(K) | 1.59 | 1.65 | 1.70 | 1.96 | 2.09 |
log(v) | 1.36 | 1.39 | 1.41 | 1.54 | 1.61 |
Plot the graph of log(K)=2log(v)+log(21m) and draw a line of best fit. From this graph, extrapolate the y-intercept:
Compare the equation of the line to Y=mX+c to identify the y-intercept:
Ylog(K)c=mX+c=2log(v)+log(21m)=log(21m)
Calculate the mass, m:
log(21m)21mm=−1.13=10−1.13=2×10−1.13=0.148…
m=0.15kg (2 d.p.)
Note: log(x) is equal to log10(x).
The case where y=abx
Given a non-linear relationship of the form y=abx, you can apply the logarithm with base 10 (or indeed any base) and use logarithm laws to obtain a linear relationship between log(y) and x:
ylog(y)=abx=log(abx)
log(y)=xlog(b)+log(a)
This forms a linear relationship of Y=mX+c where Y=log(y), X=x, m=log(b) and c=log(a).