What is Hypothesis Testing – Explained in Simple Terms?

There are 2 ways to explain Hypothesis Testing – there’s the hard and fast truth, and then there’s the slow and subtle way.

Allow me to hit you first with the hard and fast truth, before I take it slow and subtle.

For the hard and fast truth, Investopedia defines hypothesis testing as:

Hypothesis testing is used to infer the result of a hypothesis performed on sample data from a larger population. The test tells the analyst whether or not his primary hypothesis is true. Statistical analysts test a hypothesis by measuring and examining a random sample of the population being analyzed.

Now if you got lost there somewhere in that complex explanation, that’s because it is complex.

Traditional Statisticians, Data Scientists, Mathematicians, and Lean Six Sigma practicioners tend to use a lot of jargon.

But the truth is, it’s just because they know what these terms mean and it’s just part of their vocabulary when they talk to each other.

But thankfully, for the uninitiated, as well as those who are still outside of the “statistically significant” circle, there’s a slow and subtle way to explain hypothesis testing.

Allow me to break it down for you in Simple terms and give you an example.

A Hypothesis is an Assumption

A hypothesis is basically an assumption, that’s all there is to it.

For example, if you wanted to find out which mobile phone is better – the latest high-end Samsung or the equivalent iPhone, you can’t really tell if one is better than the other at this point, so all you can do is make an assumption – a hypothesis.

What is your assumption? It’s either they’re both just as good, or one is better than the other. Read on.

Null Hypothesis vs. Alternate Hypothesis

You can’t talk about Hypothesis without discussing Null versus Alternate Hypothesis.

The Null Hypothesis is the commonly accepted fact. It’s the least exciting result because it’s what people would have normally accepted as the truth if you didn’t conduct the hypothesis test. With that, you can say that the Null Hypothesis strives for natural equality.

The Alternate Hypothesis is therefore the opposite. It’s the more exciting result because it disproves the commonly accepted belief.

So in our mobile phone example, since iPhone users would say they like their phones, and Samsung users would say the same about theirs, as the person who will conduct the Hypothesis Test, your Null Hypothesis would therefore be the least exciting result – they are both just as good, so there is no difference.

Your Alternate Hypothesis, being the more exciting one, will disprove the Null Hypothesis. Therefore it’s that one mobile phone is better than the other by a significant margin.

How significant? It has to be “Statistically Significant.” And when we say “statistically” then it means it involves analysis of data.

What is Hypothesis Testing in Simple Terms?

Hypothesis Testing in simple terms is just a means for you to check if your assumption is correct, through the use of data. That’s it!

In our mobile phone example, it’s doing a comparison between iPhone and Samsung using the same criteria (user experience, picture quality, battery life, etc.) and amount of data (actual tests not just based on listed specs) to find out if they’re just as good as each other (null hypothesis) or if one is better than the other (alternate hypothesis).

That’s the end of the simple explanation on “what” hypothesis testing is.

Now do you want more details on “how” to do hypothesis testing?

Stay tuned for the upcoming Webinar on How to do Set Up and Run Hypothesis Tests.

In the meantime, what questions do you have on Hypothesis Testing? Respond with a comment. I read each and every one, and I will personally respond to you.