SHARE THE ARTICLE ON
Quadratic regression is the process of finding the equation of a parabola, that best fits your dataset.
You can identify a quadratic regression by the plotting of your scatterplots. If the scatterplots are in a shape looking like a “U” (concave up), or the scatterplot are plotted in a shape like an up-side down U like “∩” (concave down), then you can say that you have a Quadratic regression at your hand which is best fitting your data. The shape of the scatterplots are not always complete. So you might see a half U or just a 3/4th of it.
As being an extension of simple linear regression, we can say that the major difference between the both is that in linear regression, a straight line can be drawn using only two points too. But when it comes to Quadratic regression, as it is a parabolic curve, you need as much points as possible to plot a perfect curve. Due to this disadvantage of Quadratic regression, it is generally more costly than simple linear regression.
Create an actionable feedback collection process.
y = ax2 + bx + c
a = { [ Σ x2 y * Σ xx ] – [Σ xy * Σ xx2 ] } / { [ Σ xx * Σ x2 x2] – [Σ xx2 ]2 }
b = { [ Σ xy * Σ x2 x2 ] – [Σ x2y * Σ xx2 ] } / { [ Σ xx * Σ x2x 2] – [Σ xx2 ]2 }
c = [ Σ y / n ] – { b * [ Σ x / n ] } – { a * [ Σ x 2 / n ] }
Where,
a, b, c are the coefficients of Quadratic regression.
Σ x x = [ Σ x 2 ] – [ ( Σ x )2 / n ]
Σ x y = [ Σ x y ] – [ ( Σ x * Σ y ) / n ]
Σ x x2 = [ Σ x 3 ] – [ ( Σ x 2 * Σ x ) / n ]
Σ x2 y = [ Σ x 2 y] – [ ( Σ x 2 * Σ y ) / n ]
Σ x2 x2 = [ Σ x 4 ] – [ ( Σ x 2 )2 / n ]
a, b, c are the coefficients of Quadratic regression.
Let’s take an example to understand how the formula works around the give data.
Example: draw a second degree polynomial with polynomial regression for the given dataset.
x values | y values |
5 | 3 |
6 | 2 |
4 | 4 |
In our case, n=3.
x | y | x2 | x3 | x4 | xy | x2y |
5 | 3 | 25 | 125 | 625 | 15 | 75 |
6 | 2 | 36 | 216 | 1296 | 12 | 72 |
4 | 4 | 16 | 64 | 256 | 16 | 64 |
∑x | ∑y | ∑x2 | ∑x3 | ∑x4 | ∑xy | ∑x2y |
15 | 9 | 77 | 405 | 2177 | 43 | 211 |
Get market research trends guide, Online Surveys guide, Agile Market Research Guide & 5 Market research Template
Σ x x = [ Σ x 2 ] – [ ( Σ x )2 / n ]
∑xx = [77] – [(15)2 / 3]
∑xx = 77 – [225/3]
∑xx = 77 – 75
∑xx = 2
Σ x y = [ Σ x y ] – [ ( Σ x * Σ y ) / n ]
∑xy = [43] – [(15 x 9) /3]
∑xy = 43 – [135/3]
∑xy = 43 – 45
∑xy = -2
Σ x x2 = [ Σ x 3 ] – [ ( Σ x 2 * Σ x ) / n ]
∑xx2 = [405] – [(77 x 15) / 3]
∑xx2 = 405 – [1155 / 3]
∑xx2 = 405 – 385
∑xx2 = 20
Σ x2 y = [ Σ x 2 y] – [ ( Σ x 2 * Σ y ) / n ]
∑x2y = [211] – [(77 x 9) / 3]
∑x2y = 211 – [693 / 3]
∑x2y = 211 – 231
∑x2y = -20
Σ x2 x2 = [ Σ x 4 ] – [ ( Σ x 2 )2 / n ]
∑x2x2 = [2177] – [(77)2 / 3]
∑x2x2 = 2177 – [5929 / 3]
∑x2x2 = 2177 – 1976
∑x2x2 = 201
a = { [ Σ x2 y * Σ xx ] – [Σ xy * Σ xx2 ] } / { [ Σ xx * Σ x2 x2] – [Σ xx2 ]2 }
a = {[-20 * 2] – [-2 * 20]} / {[2 * 201] – [20]2}
a = {(-40) – (40)} / {(402) – (400)}
a = -80 / 2
a = -40
See Voxco survey software in action with a Free demo.
b = { [ Σ xy * Σ x2 x2 ] – [Σ x2y * Σ xx2 ] } / { [ Σ xx * Σ x2x 2] – [Σ xx2 ]2 }
b = {[-2 * 201] – [-20 * 20]} / {[2 * 201] – [20]2}
b = {[-402] – [-400]} / {(402) – (400)}
b = -2 / 2
b = -1
c = [ Σ y / n ] – { b * [ Σ x / n ] } – { a * [ Σ x 2 / n ] }
c = [9 / 3] – {-1 * (15 / 3)} – {-40 * (77 / 3)}
c = 3 – [-1 * 5] – [-40 * 25.66]
c = 3 – (-5) – (-1026.4)
c = 1034.4
y = ax2 + bx + c
y = -40x2 + (-1x) + 1034.4
y = -40x2 – x + 1034.4
Hence, the Quadratic regression equation of your parabola is y = -40x2 – x + 1034.4
Apart from this, there are various online Quadratic regression calculators that make your task easy and save all these steps and give the Quadratic regression equation straight away.
Read more
Correlation coefficient excel SHARE THE ARTICLE ON Table of Contents Correlation coefficient is used to find the relationship between two variables and how strong they
Non-leading Survey Questions SHARE THE ARTICLE ON Table of Contents What are leading and non-leading questions? Talking about leading questions, leading questions in surveys subtly
What is Employee Experience Management? SHARE THE ARTICLE ON Table of Contents Want the business gods on your side? Then this guide to employee experience
Conducting research studies can be intense! Surveys, analytics, data verification, and monitoring. One of the most rewarding parts of research is when you get to showcase your survey results to your stakeholders.
Public Opinion Polls SHARE THE ARTICLE ON Share on facebook Share on twitter Share on linkedin Table of Contents What is a public opinion poll?
Skill Gap Analysis: Definition, Steps and Benefits SHARE THE ARTICLE ON Table of Contents In a rapidly evolving job market, the need for talented professionals
English
French
