Simulations

Download: G6-Simulations_init.xlsx

Introduction

Many statistical studies use random number generators (RNGs) to simulate the behavior of random variables. These RNGs can be used to emulate the behavior of populations that follow specific probability distributions, such as normal and lognormal distributions. Both Excel and CLS-Stats have RNGs that can be used to perform simulations. This screencast is an introduction to these RNGs.

Learning Goals

• Understand how random number generators work and what RNGs are available in Excel and in the CLS-Stats library
• Understand what a volatile function is in Excel and how to manage them
• Learn to simulate random variables that follow specific distributions in Excel
• Learn how to generate reproducible and nonvolatile random numbers using the CLS-Stats library
• Learn how to use the rerror and rerrorV CLS-Stats functions to simulate random variables with specific skew and kurtosis

Tasks

• Generate 25 random numbers with RAND and RANDARRAY (and pledge to always use RANDARRAY over RAND in the future)
• Randomize a list using RANDARRAY and runif
• Draw 25 random samples from a normal distribution using RANDARRAY
• Control volatile functions by turning automatic calculations on and off
• Use CLS-Stats functions runif and rnorm and control the volatility and reproducibility of the result
• Use rerror to simulate a random variable with specific skew and excess kurtosis
• Compare the behavior (and use cases) of rerror and rerrorV
• Note some caveats for seed strings

Once you are finished with the screencast your spreadsheet should look like the image below.