Question

What is the equation of the best-fitting regression line? Find this in two ways. First, using the regression analysis part of the output, and then using the descriptive statistics part of the output.

Answer 1

Step-by-step explanation:

Regression analysis is used to infer about the relationship between two or more variables.

The line of best fit is a straight line representing the regression equation on a scatter plot. The may pass through either some point or all points or none of the points.

Method 1:

Using regression analysis the line of best fit is: $y=\alpha +\beta x+e$

Here α = intercept, β = slope and e = error.

The formula to compute the intercept is:

$\alpha =\bar y-\beta \bar x$

Here $\bar y$ and $\bar x$ are mean of the y and x values respectively.

$\bar y=\frac{\sum y_{i}}{n} \\\bar x=\frac{\sum x_{i}}{n}$

The formula to compute the slope is:

$\beta =\frac{\sum (x-\bar x)(y - \bar y)}{\sum (x=\bar x)^{2}}$

And the formula to compute the error is:

$e=y-\alpha -\beta x$

Method 2:

The regression line can be determined using the descriptive statistics mean, standard deviation and correlation.

The equation of the line of best fit is:

$(y-\bar y)=r\frac{\sigma_{x}}{\sigma_{y}} (x-\bar x)$

Here r = correlation coefficient = $r=\frac{Cov (x,y)}{\sqrt{\sigma^{2}_{x}\sigma^{2}_{y}} }$

$\sigma_{x}$ and $\sigma_{y}$ are standard deviation of x and y respectively.

$\sigma_{x}=\frac{1}{n}\sum (x-\bar x)^{2} \\\sigma_{y}=\frac{1}{n}\sum (y-\bar y)^{2}$