### Example:

### Detecting Outlier in Inflation Data

In our simple regression example we regressed the actual change in the Harmonised Index of Consumer Prices (HICP) on the predicted change in the HICP, in order to know whether the forecasts are unbiased. We used mean forecast made in the March quarter of a year from 2002 through 2014.

Suppose we modify the sample period and choose a period of 2000 through 2013 (n = 14, shown as default values in our calculator).

Determine whether the data set has an outlier at the 0.05 level of significance.

**Grubbs' Outlier Test**

**H _{0} : No outliers in the data VS H_{a} : Atleast 1 outlier in the data**

**H _{0} : Minimum value is not an outlier VS H_{a} : Minimum value is an outlier **

**H _{0} : Maximum value is not an outlier VS H_{a} : Maximum value is an outlier**

**Source : European Central Bank (ECB).**