Quantico: Hypothesis Testing

With the Quantico Shiny App, you can run hypothesis testing and generate output to help you with decision making and assumption checking. Currently, there are six methods available to the user, which includes:

1. Normality Testing
2. Correlation Testing
3. 1-Sample T-Test
4. 2-Sample T-Test
5. F-Test
6. Chi-Squared Test

App Usage

The user goes to the sidebar and clicks on Inference. Then they will see the menu of options for that category of tasks (as seen in the image below). Next, the user clicks on the method of choice and a modal pops up, prompting the user to select the necessary inputs. Next, the user will click the Run button in the top right of the modal. Lastly, the user goes to the Main Panel tab named Inference 1 to generate the output report, which includes metrics and visualizations related to the specific test they ran.

Normality Testing

The normality testing will run six different tests, including:

1. Anderson-Darling
2. Cramer-Von Mises
3. Kolmogorov Smirnov Test
4. Shapiro Test
5. Jarque-Bera Test
6. Agostino Test

You can select sample sizes and bootstrapping if you want, otherwise, as much data as possible will be passed to each test and ran once. The output includes the metrics associated with each test, a Radar Plot of p-values for each test (one for each variable), along with probability plots for each variables you pass to the procedure.

Correlation Testing

The correlation testing internally runs the correlation function from the correlation package in R. The output is plentiful and adapts to variable types you pass along for testing. Output includes a table of correlation metrics, a correlogram plot, a parallel plot, and a trend plot with standardized values of the variables for inspecting correlation over time (if you pass a date variable to the procedure).

One Sample T-Test

The One Sample T-Test generates statistical metrics, a density plot, and a normal probability plot.

Two Sample T-Test

The Two Sample T-Test generates statistical metrics, a box plot of the two variables, density plots, and normal probability plots.

F-Test

The F-Test generates statistical metrics, a box plot of the two variables, density plots, and normal probability plots.

Chi-Squared-Test

The Chi-Squared-Test generates statistical metrics, a heatmap of the observed values, and a heatmap of the residuals of observed minus expected.