Praat Statistics

Utility scripts to do statistics with Praat.

Download the scripts of this page via GitHubarrow-up-right.

Check out Praat script guide guidance for the tips of using Praat scripts.

demo_MDSexample

Screenshot of the Praat Picture window opened up after running the script.

This script combines MDS visualisation with the Shepard diagram to evaluate the mapping quality of dissimilarity data into a lower-dimensional space.

Simply run the script and the Praat Picture window will pop up to show the example.

function_drawBarPlot

Parameter specification window after running this script. The user can specify the name of the factor, dependent, min and max.

Creates a visually customisable bar plot showing means of the dependent variable across factor levels, with optional whiskers for variability and annotations for sample sizes.

Select a Table object and run the script. The output will be a a box plot shown in a Praat Picture window.

function_drawRegressionLine

Parameter specification window after running this script.

This script plots a regression line for selected data from a Table object.

Simply select a Table object and run this script to plot the line, a Praat Picture window will be opened up.

function_getZscore

Parameter window after running the script. The user can specify which group and which variable are interested for calculating the z score.

This script calculates z-scores for a specified variable within groups in a selected Table object.

Select a Table object before runnign this script, you'll get a display of z score after running it.

function_reportConfidenceInterval

Parameter specification window after running the script. The user can specify the variable of interest and the range in percent.

This function returns the confidence interval of a data column in a Table based on the t-distribution (i.e. when the population mean is unknown).

The table includes:

  • NN: number of observations.

  • df\text{df}: degrees of freedom (N-1).

  • mm: sample mean.

  • sd\text{sd}: sample standard deviation.

  • CI\text{CI}: absolute confidence interval.

  • lowerLimit: mCIm-\text{CI}.

  • upperLimit: m+CIm+\text{CI}.

Select a table containing a column with numeric data (variable). Specific variable can be chosen. The output is a new Table containing the above values.

Requirement: The Table containing the variable to be analysed must be selected.

Method: CI\text{CI} is calculated by the formula: CI=t×(sd/N)\text{CI}=\mid t\times (\text{sd} / \sqrt{N})\mid.

Last updated