Uncertainties in Data Points

In principle, there are two main methods for estimation of uncertainties in experimentally obtained data points. The first and simplest is the repetitive method. All experiments and subsequent evaluations should be made in at least triplicate. This will give a basic estimate of the variance of the adopted approach. Any less than triplicate will not leave enough degrees of freedom to calculate the standard deviation of the sample space. The standard deviation for each experimental point is calculated from the sum of the Sy values (uncertainties related to systematic errors - see Chapter 4) relating to each point as shown in Eq. (4.1). It should be noted that it is highly improbable that all points in a data series will have the same relative uncertainty although, unfortunately, this is very often the case for reported uncertainties in data points (if indeed such imcertainties are reported). 

The second and more elaborate method for estimating the imcertainty of data points is a thorough investigation of each step in the experimental process. These steps are then illustrated, for example, using a fish bone structure using a table with corresponding uncertainties. An example of such a fish bone analysis for the determination of a distribution ratio in a liquid-liquid extraction experiment using an acidic extractant and a radioactive metal tracer for concentration 

One advantage with this more cumbersome method compared to the replicate method to obtain data point uncertainty is that at any subsequent point in time, the uncertainties can be revisited and re-evaluated. In many cases, some of the 

---