Hi all,

I am conducting a single-case experimental design and using the split-middle method of trend estimation to visually analyse the data. I have the mid-date and mid-rate for each condition and thus know the two coordinates [i.e. (x,y) & (x,y)] through which I want my trend estimation line to pass through on my Graphpad graph. However, I am having trouble accurately graphing the new trend line on top of the existing data.

I have tried entering the two coordinates in a separate spreadsheet then fitting a non-linear regression by constraining the y-intercept and the slope (based on linear regression of the two data coordinates) but I find that the outcome is not accurate.

Any suggestions as to how I can create a trendline over existing data using two known coordinates would be much appreciated.

Kind regards,
Alicia