Is 0 2 greater than 0 05

What can you say if your P-value is greater than 0.05?

P-values ​​are often misinterpreted, which causes many problems. I'm not going to rehash these issues here, as my colleague Jim Frost has already detailed the issues involved, but the fact remains that p-value remains one of the most widely used tools to determine whether a result is statistically significant is, will be.

You know the old saw about "lies, damn lies and statistics", right? It sounds right, because statistics really are the maximum of interpretation and presentation because they are mathematics. This means that we citizens who we analyze data, with all our weaknesses and flaws, have the ability to obscure and obscure the way results are reported.

While I generally want to believe that people want to be honest and objective - especially smart people who research and analyze data that affects other people's lives - here is 500 pieces of evidence that contradict that belief.

We'll come back to that in a minute. But first, a quick review ...

What is a P-value and how do I interpret it?

Most people first encounter P-values ​​when we perform simple hypothesis tests, although they are part of several more sophisticated methods. Let's use the Minitab statistics software to try a quick check of how it is working (if you want to join in and don't have Minitab, the full package is available free for 30 days). We compare the fuel consumption for two different types of stoves to see if there is a difference between their means.

Go to File> Open Worksheet and click the "Browse Minitab Sample Data Folder" button. Open the sample data set named “Furnace.mtw” and select Statistics> Basic Statistics> 2 Sample t… from the menu. Inside the panel, enter “BTU.In” for samples and “Attenuator” for sample IDs.

Press OK and Minitab will return the output below with the p-value highlighted.

Most analyzes use an alpha value of 0.05 because the threshold for significance is too high. If the p-value is less than 0.05, we reject the null hypothesis that there is no difference between the means and conclude that there is a large difference. If the p-value is greater than 0.05, we cannot conclude that there is much of a difference.

It's pretty easy, isn't it? Below 0.05, significant. Over 0.05, not significant.

"Missed by so much!"

In the example above, the result is clear: a p-value of 0.7 is so much above 0.05 that you simply cannot have any illusions about the results. But what if your p-value is really on the verge of 0.05?

For example, what if you had a p-value of 0.06?

It is not significant.

It is not significant. Okay, what about 0.055?

It is not significant.

How about 0.051?

This is still not statistically significant, and data analysts should not try to pretend otherwise. A p-value is not a negotiation: if p> 0.05, the results are not significant. Period.

So what should I say if I get a p-value greater than 0.05?

How about saying this "The results were not statistically significant." If that's what the information tells you, there is nothing wrong in saying that.

It doesn't matter how thin you cut it, it's bullshit anyway.

Which brings me back to the blog post I referred to at the beginning. Thu it offers reading, but the bottom line is that the author has cataloged 500 alternative ways in which academic journalists have obscured their findings (or lack thereof) with language.

As a language student, I confess that I find the list fascinating ... but also annoying. She is incorrect: These contributors are educated people who certainly understand A) what a p-value above 0.05 means, and B) that manipulating words to fuse that result is deliberately deceptive. Or, to put it in less soft words, it's a bloody lie.

Nonetheless, this is a common occurrence.

Here are just a few of my favorites among the five hundred alternative ways that people have reported who reported results that were insignificant amid the p-values ​​that these creative interpretations applied to:

Some trend towards significance (p = 0.08)

approached the limit of significance (p = 0.07)

On the verge of statistical significance