Skip to contents

Run an Independent-Samples t-Test with Automated Narrative Reporting

Source: R/easy_ttest.R
easy_ttest.Rd

Executes a two-sample (or one-sample) t-test using stats::t.test(), extracts key metrics via broom, and generates a plain-language narrative via the Narrative Generator Module.

Usage

easy_ttest(
  x,
  y = NULL,
  data = NULL,
  mu = 0,
  var.equal = TRUE,
  conf.level = 0.95,
  alpha = 0.05
)

Arguments

x

A numeric vector, OR a formula of the form outcome ~ group when data is provided.

y

A numeric vector (second group) when x is not a formula. Ignored when x is a formula.

data

A data frame. Required when x is a formula.

mu

Null hypothesis value for the mean (one-sample test). Default 0.

var.equal

Logical; assume equal variances? Default TRUE (classical pooled-variance t-test). Set to FALSE to use Welch's t-test instead.

conf.level

Confidence level. Default 0.95.

alpha

Significance threshold for narrative. Default 0.05.

Value

An object of class "easystat_result" with:

test_type

Character: "ttest"

formula_str

Description of the comparison

raw_model

The raw htest object

coefficients_table

Group means and confidence interval

model_fit_table

t-statistic, df, and p-value

explanation

Plain-language narrative string

Examples

result <- easy_ttest(mpg ~ am, data = mtcars)
print(result)
#> 
#> ================================================================================
#>  EasyStat Result :: TTEST
#> ================================================================================
#> 
#> TABLE 1 — MAIN RESULTS
#> --------------------------------------------------------------------------------
#>                              Metric   Value
#>                            Mean — 0 17.1474
#>                            Mean — 1 24.3923
#>  Mean Difference (larger − smaller)  7.2449
#>                      95% CI (lower)  3.6415
#>                      95% CI (upper) 10.8484
#>                               n — 0 19.0000
#>                               n — 1 13.0000
#> 
#> TABLE 2 — MODEL FIT / SUMMARY
#> --------------------------------------------------------------------------------
#>                                        Metric                   Value
#>                                 |t-statistic|                  4.1061
#>                     Degrees of Freedom (used)                      30
#>  Degrees of Freedom (equal variance, n1+n2-2)                      30
#>                           Variance Assumption Equal variances assumed
#>                                       p-value                  0.0003
#> 
#> ================================================================================
#>  PLAIN-LANGUAGE INTERPRETATION
#> ================================================================================
#> 
#> INDEPENDENT-SAMPLES t-TEST Comparison: mpg ~ am
#> 
#> An independent-samples t-test revealed a highly statistically significant (p
#>   = 0.0003) difference between the two groups (t(30) = -4.106). The mean for
#>   '0' was 17.1474 and the mean for '1' was 24.3923. The 95% confidence
#>   interval for the difference in means ranged from 3.6415 to 10.8484. These
#>   results provide statistically significant evidence that '0' and '1' differ
#>   meaningfully on the measured variable.
#> 
#> ================================================================================
#>