## Method

Many Standard R outputs contain a default output which is a table, similar to the two below.

### Coefficient tables

Statistical models that estimate coefficients are often represented in Standard R with a table such as the one below. Key aspects of this table are:

- The estimated parameters, which are referred to as
`Estimate`s in the example below, are color-coded where the corresponding*p*-value is less than 0.05. If the coefficient is positive they are colored blue. If negative, they are colored red. - The
*t-*or*z*-statistics are colored from coral (a pink color) through to blue, where:- Values less than -5 are shown in the strong coral (e.g., as shown in the
`sleepy`row below. - Values of 0 are shown with white background.
- Values of 5 or more are shown in blue.
- Values between -5 and 5 are interpolated based on the
*z*and*t*statistics.

- Values less than -5 are shown in the strong coral (e.g., as shown in the
*p-*Values are shown in the column marked*p*. They are bold where*p*is less than 0.05.

### Tables of means

The table below shows means by group, where the rows represent variables, and the initial columns represent different sub-groups in the population. As with the *Coefficient tables*, the shading is determined by the *t*-statistics (computed as differences from the row means). However, whereas in the table above the *t*-statistics themselves were shaded, in this example the means are shaded according to the *t*-statistics, and the *t*-statistics themselves are not displayed. See Machine Learning - Linear Discriminant Analysis for a more detailed discussion of this example.

## Next

## Comments

0 comments

Article is closed for comments.