Validated Sampling Method for InspectionsAuditsInventories Gary Robinson Palm

Validated Sampling Method for Inspections/Audits/Inventories Gary Robinson Palm Beach County Sheriff’s Office Professional Compliance Bureau Staff Inspections/Accreditation

Objectives • Provide a valid sampling method to use in audits, inspections and inventories. • Provide the tools to “do the math” without doing the math! • Present the methods to use the tools. • Provide examples that can be taken back, adapted and used in your agencies.

Why use this method? • Eliminates questions about validity of the method used to select the samples/items for inspection/inventory. • Uses available, accepted and validated websites for the selection process. • Adapts to any size/complexity of your organization. • Allows for professional judgement in selecting samples in unique or exigent circumstances. • The sampling method complies with the GAO Generally Accepted Government Auditing Standards (GAGAS; “the yellow book”).

Now for the boring stuff… • Basic statistics – sampling occurs on the (infamous!) “bell-shaped curve” “A bell curve is another name for a normal distribution curve (sometimes just shortened to “normal curve”) or Gaussian distribution. The name comes from the fact it looks bellshaped. The term “bell curve” is usually used in the social sciences; in statistics, it’s called a normal distribution and in physics, it’s called a Gaussian distribution. However, they all refer to exactly the same thing: a probability distribution that has certain characteristics, including the fact it’s shaped like a bell. “ Source: https: //www. statisticshowto. datasciencecentral. com/bell-curve/

So, how do I get a sample out of that bell? • Two ways to “ring” it – a “two-tail” sample or a “one-tail” sample. • Two-tail involves sampling both sides of the curve; one-tail only samples one side. • Don’t worry: it’s a math exercise someone smarter than us has already done for us! • The main differences are the two-tail will result in a much larger sample; one-tail will return, at a maximum, 96 samples (it’s a “math-thing”)

Which one should I use? • Here’s where that “professional judgement” comes into play. • If you’re a small agency with a single location, relatively few items, limited staff, and no concerns with your evidence, a one-tail sample may be sufficient. • For larger agencies or if concerns are present, the two-tail will provide a deeper sample.

Source: Law Enforcement Inspections and Auditing courses; LD Consulting LLC

So, how do I get those smarter people to do the math for me and my stuff? • Simple! It’s all out there on the “inter-web” for you! But you have a little work to do first – 1. Determine what you’re going to use for your population (the total number of things you will sample from). These can be: - Case numbers - Packages (submissions; log numbers; etc. ) - Any other group of items you wish to audit

2. Determine what your “strata” are (that’s the places where you’ve got stuffed) Bet you’re asking why? Well, this will allow you the broadest sample and to meet the “extra security measures” element of the standards. The basic places are usually: - General property - Evidence - Narcotics - Firearms/weapons - High-value (money/jewelry/negotiable instruments) 3. Determine the population of stuff in each strata.

Okay, I’ve got all that. Now what? • Now you need the tools: – Working spreadsheet – Sample Size Calculator – Randomizer – Population spreadsheet of all the stuff in each strata

Altogether, now! • Altogether • BIG Altogether (well, after a bunch of work) – Sample Inventory Sheet • The final report • Policy, you ask? Of course! Wait for it…

Shameless Plug! September 10 -12, 2019, Part 2, LEIA-201 (needed for LEIAC certification) is October 21 -24, 2019 at PBSO Special thanks to Dan Olson (Lt. Dan) of LD Consulting LLC and the Sanford Police Department for permission to share some of the material and spreadsheets provided. For a copy of this presentation and examples, email me at: robinsong@pbso. org