Prediction of a Specific Observation

Many times researchers are not interested in an expected (mean) value or mean value CL They instead want an interval for a specific y, value corresponding to a specific x,. The process for this is very similar to that for the expected (mean) value procedure, but the Cl for a single, new y, value results in a wider Cl than does predicting for an average y, value. The formula for a specific y,- value is 

the researcher can expect the value y, Gogio microorganisms) to be contained within the 5.13 to 5.90 logio interval at a 15 sec exposure at a 95% confidence level. This does not mean that there is a 95% chance of the value being within the CL It means that, if the experimental procedure was conducted 100 times, approximately 95 times out of 100, the value would lie within this interval. Again, this is a prediction interval of one y, value oti rme Xi value. 

---