Validation is a vital concept when it comes to proving what things do.  It is a term commonly used by engineers, scientist, physicians and other people working in evidence based professions.  I usually need to explain how I have validated a new method or tool to my colleagues or customers, because it is an important part of proving my work does what I designed it for.  In addition, like other engineers and scientists, I will often ask how a new method or tool has been validated.  If there is no validation evidence I cannot use it for my work unless I validate it myself.

So what does ‘validation’ mean, and why is it so important to scientists, engineers and physicians?

What is validation?

In simple terms, validation is the process of demonstrating that something does what it is supposed to do.

Scientific measurement is mostly about collecting evidence.  In open ended research this is often to define a phenomenon that is not understood.  For example, cancer researchers take measurements to learn how cancer behaves and what we can do to change it.  They also measure the effects of their experimental treatments to see if the are working as expected and treating the cancer.

This later type of measurement is a type of validation.

In business, validation is often about proving that your design satisfies customers.  In engineering, it is usually about proving that your design does what it is supposed to do, safety and within any legal requirements.  In medicine, it may be proving that a therapy effectively treats the conditions it is targeting, or that a test accurately detects the disease it is testing for.

A simple example of validation

Lets imagine an new invention that is going to change the world and disrupt the measurement of people.  Imagine that I have invented a device that clips on your little finger and measures your height, weight and age.  This is going to revolutionise measurement!  Athletes can be categorised quickly at a desk, patients don’t have to be conscious and standing to accurately work our how much medicine they need, and parents don’t have to put marks on their wall to settle arguments about who is taller.

As fantastic as this imaginary invention is.  There would be some serious consequences if it was not accurate.  Patients may receive incorrect doses of medicine, gold medals may be unfairly won, and arguments over who is taller may never end.

So before I start producing and selling my device, I had better prove that I have designed something that accurately measures height, weight and age.  How?  In this case it is easy because I already have access to standard accurate ways to measure these things.  We have measuring tapes, bathroom scales, and dates of birth.  All I need to do is take measurements with my device and the standard methods, and compare them. 

(Actually it is not quite that simple. It is important that I use statistics to work out how many measurements need to be taken during my validation experiment.  However, that discussion will require an article of it’s own.)

For the sake of this example, if my design works, then my prototype device will give the same measurements as the measuring tape, the bathroom scale, and asking someone’s date of birth.  If that is true then I have proved that my design works by validating it against standard ways of measuring people.  This also means that I have evidence to show my customers that the measurements from my device are at least as accurate as those from a tape measure, a bathroom scale, and asking someone their age.

Why is validation important?

Now this imaginary example may seem trivial, but think of some examples of your own.  How important is it that a builder’s laser level is accurate?  Is it important that my brakes slow down my car when I press the pedal?  Does my computer’s encryption actually stop hackers?  How important is it that my banks computer systems transfer the correct amount of money into my account?

Now you what validation is, and why engineers, scientists, physicians, and many others take it so seriously.  Next time you see a great new technology or other solution that seems ‘too good’, you know what you should expect the designer to have done to prove that it works.  Validate it.