Calculating a slope by the function and graph on a semi-log graph

[ARGUANI]

Hello,

I am trying to calculate the slope of the following data:
Y X
0.00061000/0.00264445
0.00114000/0.00263054
0.00222000/0.00260994
0.00412000/0.00258967
0.00758000/0.00256970

On a semi-log graph (Y-log) by the slope function: Slope = (log(Y2)-log(Y1))/(X2-X1)
And, also by the graph. However, I am finding unmatched values! (-33074 by the graph and -14639 by the function)

Any help will be appreciate it.

Thanks

[MrShorty]

I note that, if I plot and regress based on common log (log₁₀), then I get m=-14364. If I regress on natural log (ln() or log_e), then I get m=-33075. I am guessing that in one case Excel is using common log and the other is using natural log. Whichever one you decide is "the correct" way to express it, make sure you are clear which base you are using.

[ARGUANI]

Dear MrShorty,

Thank you for your quick replay and clarification. I can know what the base of the log in the function, but how can I check that in the plot?

Thank you for your help again.

[ARGUANI]

Please find attached the pic of the graph. The base of the log used in the plot is 10, but the slope in the equ. is -33074 not 14364!
Any advise?

Thanks

[MrShorty]

The given equation is of the form y=Ae^(Bx). Since e is the base for natural logs, that indicates that Excel did the regression using natural logs. If it had used common logs, the equation would have been y=A*10^(Bx).

[ARGUANI]

MrShort,

Thanks a lot for your help.