You are on page 1of 2

Problem Statement Regression

Student Xi Yi xi(Xi-mean) yi(Yi-mean) x2 y2 xy


1 65 70 -12 -8 144 64 96
2 95 85 18 7 324 49 126
3 45 95 -32 17 1024 289 -544
4 85 65 8 -13 64 169 -104
5 68 75 -9 -3 81 9 27
sum 385 390 1637 580 -399
mean 77 78

What linear regression equation best predicts the statistics performance, based test?
Calculation:
The given equation is ŷ = b0 + b1x
b1 = Σ [ (xi - x)(yi - y) ] / Σ [ (xi - x)2]
b1 = -399/1637
b1 = -0.243
Now, For b0
b0 = y - b1 * x
b0 = 78 - (-0.243)(77)
b0 = 96.711
So, the equation will be now,

ŷ = 96.711 - 0.243x

If a student made an 80 on the test, what grade would we expect her to make in online course?
ŷ = 96.711 - 0.243x
ŷ = 96.711 - 0.243*80

ŷ = 77.271
How well does the regression equation fit the data?
Coefficient of determination:
R2 = { ( 1 / N ) * Σ [ (xi - x) * (yi - y) ] / (σx * σy ) }2
For σx:
σx = sqrt [ Σ x2 / N ]
σx = sqrt(1637/5) = sqrt(327.4)
σx =18.094
For σy:
σy = sqrt [ Σ y2 / N]
σy = sqrt(580/5) = sqrt(116)
σy = 10.770
Now, by putting all values.
R2 = { ( 1 / 5 ) * -339 / (18.094 * 10.770 )}2
R2 = (-67.8/194.872)2
R2 = 0.121
A coefficient of determination equal to 0.12 indicates that about 12% of the variation in statistics grade.

You might also like