A baseline signifies a normal, expected value and makes changes from the norm obvious and calculable. Baselines can be used for anything from health concerns such as heart rate, cholesterol or weight, to financial matters such as income and expenses. Essentially, a baseline calculates as an average taken when conditions are normal and not influenced by unusual events. For example, you would measure your baseline heart rate at rest, rather than after running five miles when your heart rate is unusually high.

#### Step

Maintain a record of measurements with as many data points as feasible. The accuracy of your baseline increases as the number of data points increases. In general, the more data you collect, the greater the accuracy achieved.

#### Step

Average the data entries by totaling the numbers and dividing the sum by the number of entries. The resulting figure is your baseline average. As an example, the data 100, 150 and 200 would be averaged as (100+150+200) / 3, which equals 150.

## Video of the Day

#### Step

Obtain a measure of variability within your data by calculating the standard deviation. For each individual sample measurement, subtract it from the mean and square the result. If the result is negative, squaring it will make it positive. Add all these squared numbers together and divide the sum by the number of samples minus one. Finally, calculate the square root of the number. In the prior example, the average is 150, so the standard deviation would be calculated as the square root of [[(150-150)^2+(150-100)^2+(150-200)^2]/(3-1)], which equals 50.

#### Step

Determine the standard error. The standard error allows construction of a confidence interval around your average. The confidence interval gives a range in which some percentage -- usually 95 percent -- of future values will fall. The standard error is calculated by taking the standard deviation and dividing it by the square root of the number of data points. In the prior example, the standard deviation was 50 with 3 data points, so the standard error would be 50 / squareroot(3), which equals 28.9.

#### Step

Multiply your standard error by two. Add and subtract this number from your mean to get the high and low values of a 95 percent confidence interval. Future measurements that fall within this range are not significantly different than your baseline. Future measurements that fall outside this range denote a significant change from your baseline.

In the prior example, the average was 150 with a standard error of 28.9. 28.9 multiplied by 2 equals 57.8. Your baseline will read "150 plus or minus 57.8." As 150 plus 57.8 equals 207.8, and 150 minus 57.8 equals 92.2, the baseline results in a range of 92.2 to 207.8. Thus, any measurement between these two figures is not significantly different from the baseline, because the range takes into account the variability of the data.