Is "abnormality" a correct word in this context?

[cigogne]

Hi,

In a simulated driving, the researchers have examined the number of vehicle collisions with pedestrians as a variable. I want to say that this number is not normal. Is it right to use the word "abnormality"?

"Because of the abnormality of the number of collisions with pedestrians, the Kruskal-Wallis statistical test was used to determine the differences in the values of this variable among the three age groups, where the obtained P-value was 0.536."

Thanks!

 

[Scrawny goat]

No. Do you mean that the data are not in a normal distribution?

 

[lingobingo]

In order to understand "abnormality" the reader needs to know what "normal" is. If they certainly would, then it's fine. If not, it might be better to say, for example, because of the unexpectedly large number of collisions……

 

[cigogne]

There are different variables. Some of them are normal and some are not, depending on the obtained p-value.

 

[Scrawny goat]

There are different variables. Some of them are normal and some are not, depending on the obtained p-value.

This sounds like a very technical use of the word 'normal'. I'm not familiar with it.

 

[atokad]

No, the word "abnormality" is not used in this context. You could say "non-normality," but I think it would be better to rephrase it: "Because the number of collisions does not follow a normal distribution, the Kruskal-Wallis..."

 

[ravonseed]

I would use another word for abnormality
try: deviation / irregularity / divergence / uncommonness

 

[atokad]

I would use another word for abnormality
try: deviation / irregularity / divergence / uncommonness

cigogne is not talking about the ordinary meaning of "normal"; this is talking about a technical term from statistics. A normal distribution is a Gaussian distribution, aka a bell curve.

 

[Andygc]

this is talking about a technical term from statistics. A normal distribution is a Gaussian distribution, aka a bell curve.

But it isn't (talking about a technical term).

There are different variables. Some of them are normal and some are not, depending on the obtained p-value.

The p-value doesn't tell you if the distribution is normal, it tells you the probability of the findings being due to chance.

cigone appears to be using "normal" wrongly.

 

[atokad]

Sure, I know what a p-value is, and the sentence "Some of them are normal and some are not, depending on the obtained p-value" doesn't make a lot of sense to me, but nevertheless, I think cigogne is talking about normal in the Gaussian sense. This is based on skimming the Wikipedia article about the Kruskal-Wallis test, which explains:
"Since it is a non-parametric method, the Kruskal–Wallis test does not assume a normal distribution of the residuals, unlike the analogous one-way analysis of variance."

But hey, I could be wrong. Only cigogne knows for sure.

 

[Andygc]

Only cigogne knows for sure

and needs to tell us what he means by "normal".

In a simulated driving, the researchers have examined the number of vehicle collisions with pedestrians as a variable. I want to say that this number is not normal.

We can have a normal distribution, but what is a "normal" number?