The unusual observation table displays observations that have a disproportionate impact on the regression model. These points are important to identify because they can produce misleading results. For example, a significant coefficient may appear to be nonsignificant.

There are two types of influential observations.

·    Large residuals (R): These points are extreme in the y-direction relative to the fitted regression line. Standardized residuals with absolute values greater than 2 are marked as large.

·    Leverage points (X): These points are extreme in the x-direction. If the leverage value is greater than 3 * number of model terms/number of observations, it is marked as a leverage point.

These cases do not follow the proposed regression equation well. However, it is expected that you will have some unusual observations. For example, based on the criteria for large residuals, you would expect roughly 5% of your observations to be flagged as having a large residual.

For influential observations, you should investigate whether the data were recorded correctly, and whether the data collection process was affected by any other factors. To determine the extent of influence, you can fit the model with and without an influential observation and compare the coefficients, p-values, R2, and other model summary values.

If the analysis indicates that there are many unusual observations, the model will most likely exhibit a significant lack-of-fit. That is, the model does not adequately describe the relationship between the factors and the response variable.

For the catalytic reaction data, one observation, number 2, has a standardized residual of - 2.02. Because there are 14 observations in this study, one unusual observation is expected. You should investigate this observation.