I’m sure I probably learned this in high school but… whatever.

Say we have two groups of people, group A and group B.

In group A 30% of people have a certain characteristic; obesity, for this example.

In group B 45% of people are obese.

There is a difference between these two groups of (45-30=) 15%.

From another perspective however there is a 50% difference between the two groups (30+(0.5×30)=45

Is the second perspective a valid one to use when statistically analyzing the differences in obesity between the two groups?

If it is valid, how do we distinguish between the first perspective (a 15% difference) and the second (50% difference)? How do we refer to each of these numbers (eg one is percentage, the other is percentage points – I don’t know if thats correct, just giving an example). When is it appropriate to reference the first and when is it appropriate to reference the second in terms of what one is trying to demonstrate with the statistics?

Also, what about the (45+(-0.3333×45)=30%, giving a difference between the two of -33%? How does this fit in with all the rest?

Sorry, I’ve basic ally forgotten all my stats.

esselte said:

I’m sure I probably learned this in high school but… whatever.

Say we have two groups of people, group A and group B.

In group A 30% of people have a certain characteristic; obesity, for this example.

In group B 45% of people are obese.

There is a difference between these two groups of (45-30=) 15%.

From another perspective however there is a 50% difference between the two groups (30+(0.5×30)=45

Is the second perspective a valid one to use when statistically analyzing the differences in obesity between the two groups?

If it is valid, how do we distinguish between the first perspective (a 15% difference) and the second (50% difference)? How do we refer to each of these numbers (eg one is percentage, the other is percentage points – I don’t know if thats correct, just giving an example). When is it appropriate to reference the first and when is it appropriate to reference the second in terms of what one is trying to demonstrate with the statistics?

Also, what about the (45+(-0.3333×45)=30%, giving a difference between the two of -33%? How does this fit in with all the rest?

Depends on context and what you’re looking for/at/testing, both are valid.

I don’t know the correct terminology, but if you are discussing with non-statisticians they probably won’t either, so I’d suggest just clearly stating the context.

sibeen said:

Sorry, I’ve basic ally forgotten all my stats.

don’t be sorry, your brain thanks you for it

Arts said:

sibeen said:

Sorry, I’ve basic ally forgotten all my stats.

don’t be sorry, your brain thanks you for it

There’s a quite high probability that you’re correct.

esselte said:

I’m sure I probably learned this in high school but… whatever.

Say we have two groups of people, group A and group B.

In group A 30% of people have a certain characteristic; obesity, for this example.

In group B 45% of people are obese.

There is a difference between these two groups of (45-30=) 15%.

From another perspective however there is a 50% difference between the two groups (30+(0.5×30)=45

Is the second perspective a valid one to use when statistically analyzing the differences in obesity between the two groups?

If it is valid, how do we distinguish between the first perspective (a 15% difference) and the second (50% difference)? How do we refer to each of these numbers (eg one is percentage, the other is percentage points – I don’t know if thats correct, just giving an example). When is it appropriate to reference the first and when is it appropriate to reference the second in terms of what one is trying to demonstrate with the statistics?

Also, what about the (45+(-0.3333×45)=30%, giving a difference between the two of -33%? How does this fit in with all the rest?

The obvious first question is how many people are in groups A and B.

If either the number of people in group A or B is small, or both, then saying 15% difference or 50% difference could both be wrong, there could be no difference between the two groups. So before you say anything do an unpaired t test on the data.

However, let’s assume that there are very many people are in groups A and B. Then “The mean difference, or difference in means, measures the absolute difference between the mean value in two different groups” and is 15%. However, it also valid to say that the mean in group B is 50% larger than the mean in group A.

However, I would always replace the word “mean” in the above paragraph with “sample mean”.

ABSOLUTE and RELATIVE

see also

http://www.abc.net.au/news/2018-07-05/fact-check-flu-vaccine/9906284
https://theconversation.com/the-flu-vaccine-is-being-oversold-its-not-that-effective-97688

cken geniuses