Rounding Rules for this Class

These are the rounding rules for Dr. Gurney's Math 241 classes. In real world situations, you may be required to use different rounding rules.

  1. When computing statistics, use one more place accuracy than the maximum accuracy of the data. Two more places will be allowed but using more than two places beyond the accuracy of the data will result in penalties or REDO's. If you are familiar with significant digits, than use at most one more significant digit than the maximum used in the data that was presented.

  2. When computing fractions or probabilities, if the result has exactly one decimal place like 0.3 or 0.8, then use the result as is. Otherwise, round the result using 2 to 4 decimal places. If you understand significant digits, the more general rule is to round the result using 2 to 4 significant digits.

  3. When finding numbers that are bounds to given probabilities, as when using the TI-83/84's invNorm function, do not use more than one place beyond the accuracy of the mean given in the problem. There is usually no reason to guarantee the accuracy of the given mean and standard deviation, so there is no reason to guarantee the accuracy of any number computed from these values.

  4. When computing sample sizes needed to estimate a population mean or estimate a population proportion, if there is any non-zero amount to the right of the decimal point, you must round the result UP TO THE NEXT INTEGER.

  5. If you are not computing sample sizes needed to estimate a population mean or a population proportion, you may round up or round down, or use an intermediary rounding rule like the "Keep It Even" rule. The "Keep It Even" rule says that you round to the nearest value at the the accuracy you desire; and if the number you are rounding has one more place than desired and the final digit is 5, then you always round to make the new final digit an even number.

 

 

Return to Supplementary Statistics Topics