# Calculating regression (beta and alpha) by hand

#### Almostasmartstudent

##### New Member
Hey guys,

I have to calculate the following by hand:

1. A estimates regression (ß and α) with the dependent y and independent x;
2. The correlation between y and x;
3. The residue for the respondent

N = 16
The independent variabele is a continue one with a mean of 19.0 and a SD of 3.7
The dependent variabele is also continue and has a mena of 15.0 and a SD of 2.5
De standardized slope of the regression (so beta) is 0.60
One respondent of the dataset had a y = 13.0 and x= 15.0.

I don't know where to start ):

#### fed2

##### Active Member
What do you got so far?

I think the 'standardized slope' and correlation are actually on in the same? going from unstandard to standard slope should involve multiplying by ratio of sds of oy and x. getting alpha go with alpha = y mean - beta (unstandardized) xbar { or something like that check your notes}

see how that goes.

#### Almostasmartstudent

##### New Member

I wanted to use this formula, but since x - x (15 - 15) = 0.. I can't get a slope. Becuase 0 * y / 0 = 0. So I'm already stuck at the first quesiton.

#### fed2

##### Active Member
you probably haven't learned how to divide by zero yet?

Let's try look at formula for "De standardized slope of the regression (so beta) "? Does it seem like some relationship to unstandardardized.