Sign In
When the first differences of a set of data are constant, the most appropriate model is a linear model. Think about the second differences. What do they have to be for a quadratic model to be appropriate?
It is appropriate to use a quadratic model for a set of data when the second differences are constant.
When data has equally-spaced inputs, we can analyze patterns in the differences of the output to determine what type of function can be used to model the data. Let's see two examples.
Consider a function f(x) represented by a table of values.
We see that the table has equally-spaced x-values. Moreover, the first differences are all equal to 2, and therefore they are constant. This means we can model the data with a linear function. Therefore, a quadratic model is not the most appropriate one for this set of data.
We see that the first differences are not constant. This means that a linear model is not the most appropriate one. To determine if a quadratic model is the most appropriate one, we will calculate the second differences. To do so, we need to subtract consecutive terms of the first differences.
x= 0, g(x)= 1
Calculate power
Zero Property of Multiplication
Rearrange equation
(II): Add I
(II): Remove parentheses
(II): Add and subtract terms
(II): .LHS /2.=.RHS /2.
(I): a= 1
(I): LHS-1=RHS-1
(I): LHS * (- 1)=RHS* (- 1)