# Effect measure for association between one continuous and one categorical variable

Source:`R/effect.R`

`effect_summary.Rd`

User can either use or extend these functions to configure effect calculation.

## Usage

```
diff_mean_auto(x, by, conf_level = 0.95, R = 500)
diff_mean_boot(x, by, conf_level = 0.95, R = 500)
diff_median_boot(x, by, conf_level = 0.95, R = 500)
diff_mean_student(x, by, conf_level = 0.95)
```

## Arguments

- x
numeric vector

- by
categorical vector (of exactly 2 unique levels)

- conf_level
confidence interval level

- R
number of bootstrap replication

## Functions

`diff_mean_auto()`

: (**Default**) calculate a specific "difference in means" effect based on normality (Shapiro or Anderson test) and variance homogeneity (Bartlett test)`diff_mean_boot()`

: calculate a "difference in means" effect with a bootstrapped CI using standard deviation`diff_median_boot()`

: calculate a "difference in medians" effect with a bootstrapped CI using quantiles#'`diff_mean_student()`

: calculate a "difference in means" effect using`t.test`

confidence intervals