CHIP Error Tolerance

In all current CHIP courses, the default tolerance allowed for essentially all of the numerical answers is 1 % of the correct answer. Please read on to see what correct answer refers to as far as these CHIP problems are concerned.

1. How CHIP correct answers are arrived at:

   Unless otherwise noted, treat all numerical values given in CHIP problems as if they were exact. Use the physical and other constants as provided in the tables linked from CHIP or in your texts. Do not round the values at intermediate stages. Enter at least 3 significant figures in all numerical answers you put in CHIP answer boxes.

2. 1 % error tolerance:

   1 % error tolerance means that any answer which is within 1 % deviation in either direction from the correct answer as above will be accepted as correct. In cases where the correct answer happens to be either anomalously small or even zero for some randomly generated input values in the problem, we try to allow a reasonable absolute tolerance rather than the default 1 %. Also for occasional problems which clearly demand more (or less) accuracy, this limit may be tightened or relaxed as necessary (and a statement to this effect will be provided).

3. Common mistakes:

   • If a number calculated for one part of the problem is to be used to calculate the answer for another part, be sure to use all the digits you calculated for the first number for calculating the second, even though you only need to enter each number into the answer boxes to 3 significant figures.

   • For evaluating trignometric functions, if you need to use explicitly, do not approximate it to 3.14, but rather use 3.14159 (or better). This is particularly true for operations like sin(2ft) where ft can be large.

4. Is physics about giving many significant digits?

   No! Physics is not about calculating many digits per se, but it is about analyzing the essential features of a problem, applying the underlying concepts, and obtaining quantitative results. The result in numerical form must be given only to the precision warranted by the input data. However, the CHIP exercises are providing a hypothetical environment where all input data are treated as if they were exact - primarily because the original sources of many of the CHIP problems do not consistently specify the accuracies of the input numbers and also because of the need to produce randomized input numbers. In light of this situation, we decided that CHIP exercises will mainly test your understanding of which concepts are relevant and how they apply quantitatively - rather than the specific relationships between the precision of the input data and that of the resulting predictions.