You measure the values of resistance along a wire at intervals in x,
with distances xi determined very accurately and values of resistance Ri
each with an error so that the error flags would be (+) and (-) .04 ohms.
The data are given below along with useful sums. The values of R are from the equation
R = b + m x, for the values of b and m determined in the least squares fit.
# |
xi |
Ri |
x2i |
xiRi |
R |
|
m |
ohms |
ohms |
||
1 |
10 |
1.32 |
100 |
13.2 |
1.31 |
2 |
20 |
2.30 |
400 |
46.0 |
2.27 |
3 |
30 |
3.20 |
900 |
96.0 |
3.22 |
4 |
40 |
4.19 |
1600 |
167.6 |
4.18 |
5 |
50 |
5.08 |
2500 |
254.0 |
5.13 |
6 |
60 |
6.09 |
3600 |
265.4 |
6.08 |
7 |
70 |
7.01 |
4900 |
490.7 |
7.04 |
8 |
80 |
7.99 |
6400 |
639.2 |
7.99 |
9 |
90 |
8.96 |
8100 |
806.4 |
8.95 |
10 |
100 |
9.92 |
10000 |
992.0 |
9.90 |
SUMS |
550 |
56.06 |
38500 |
3870 |
From the sums in the table,
so
then b is given by
so
b = (38500 x 56.06 - 550 x 3870)/82500 = 0.361
where the common error on each measurement has canceled in this case. Similarly m is found from:
so
m=(10 x 3870 - 550 x 56.06)/82500 =0.0954
where again the common measurement error cancels. Now we can determine the uncertainty in these fitted values.For our case of common measurement errors:
giving the result that: m=.0954 ±.0004. The error in b can be calculated from
so that b=0.36 ±.03 Ohms.
The final result is: R = 0.36 + (0.0954 x) ohms with an error in R
given by
A plot of the data with fitted line:
From the appearance of the plot, namely that the fitted line passes within all the error bars,we can infer that the measurement error is probably overestimated.