fit data to a "HORIZONTAL" line, not a ordinary st

[lizherr98]

I need to fit a set of data points through a horizontal line. Currently I have a set of 4 points (I'm working with a small population), and it gives me a R squared value of 0.2751 (showing poor correlaton), however it has a definite slope. I need to find out what the R squared be for a horizontal line, and would the R squared values be significantly different. I tried using slope=0 (y=mx + b), which makes y=b, then I maintained that number for the Y value, then I solved for the slope using my data. I ended up with the same Y values and the original X values, this gives me a horizontal line but I'm not sure if that is the correct method to solve it. Any help will be appreciated it.

Thanks.

 

[stapel]

If the data's regression line is clearly not horizontal, then why are you needing to try to force a fit to a horizontal equation?

It might help if you provided the original statement of the exercise. Thank you.

Eliz.

 

[marcmtlca]

Show us the data and some calculations.