Option (A) is correct.
Explanation: The error is the sum of all the terms she left out. With the row 1,8,28,56,70,56,28,8,1 and x=0.03:
| term left out | value |
|---|
| 8C2x2=28×0.0009 | 0.0252 |
| 8C3x3=56×0.000027 | 0.001512 |
| 8C4x4=70×0.00000081 | 0.0000567 |
| the last five terms together | less than 0.000002 |
The last five terms are 56x5≈0.00000136, 28x6≈0.00000002, and three even smaller ones, so together they are less than 0.000002.
Lower bound. All the terms are positive, so
error>0.0252+0.001512+0.0000567=0.0267687>0.026
Upper bound.
error<0.0267687+0.000002=0.0267707<0.027
So the error lies in (0.026, 0.027). (In fact (1.03)8=1.26677… and the error is 0.02677…)
Why the other options are wrong
- (B) This keeps only the first term left out, 0.0252. The next term alone, 0.001512, pushes the error above 0.026.
- (C) This takes the error to be the first left-out term alone and works it out as 2n2x2=264×0.0009=0.0288, which lands in (C). The coefficient is 8C2=28×7=28, not 282=32; and the later left-out terms must be added too: even with 32, adding 0.001512+0.0000567+… gives about 0.0304, outside (C).
- (D) This interval suits an error of just over x2=0.0009: it takes x2 plus a little for the later terms and forgets the coefficient 8C2=28. The first term left out alone is 28×0.0009=0.0252.