REGISTER NOW!! Why? Because you can't do much without having been registered!

At the moment you have limited access to view all discussions - and most importantly, you haven't joined our community. What are you waiting for? Registration is fast, simple and absolutely free so please, join Join Talking Horses here!

Reply to thread

Message

From a stats perspective, you'd probably consider using something called the 5% trimmed mean. There are certainly societies (the UK isn't the worst example, but it does exhibit some of the traits) where an outrageous amount of wealth is found in the top 1% and this can most definitely skew an arithmetic mean.

This is a frequent mistake in stats when people talk about "the average" 99 times out of 100 (couldn't resist that), they're actually talking about the arithmetic mean, which is just one measure of an average (you all know that) but it's amazing how many times we allow ourselves to get sucked into this error. You might even find the mode is the better indicator in this case

Silly example - but it's often used to illustrate this kind of thing:

How many arms has the average person got?

If you used the mean the answer would be 1.999
If you used the mode it would be 2

If you were manufacturer of shirts and decided to make products based around the average (as defined by the mean) would that be sensible?

<blockquote data-quote="Warbler" data-source="post: 2518797" data-attributes="member: 65228">From a stats perspective, you&#039;d probably consider using something called the 5% trimmed mean. There are certainly societies (the UK isn&#039;t the worst example, but it does exhibit some of the traits) where an outrageous amount of wealth is found in the top 1% and this can most definitely skew an arithmetic mean.This is a frequent mistake in stats when people talk about &quot;the average&quot; 99 times out of 100 (couldn&#039;t resist that), they&#039;re actually talking about the arithmetic mean, which is just one measure of an average (you all know that) but it&#039;s amazing how many times we allow ourselves to get sucked into this error. You might even find the mode is the better indicator in this caseSilly example - but it&#039;s often used to illustrate this kind of thing:How many arms has the average person got?If you used the mean the answer would be 1.999If you used the mode it would be 2If you were manufacturer of shirts and decided to make products based around the average (as defined by the mean) would that be sensible?</blockquote>

[QUOTE="Warbler, post: 2518797, member: 65228"] From a stats perspective, you'd probably consider using something called the 5% trimmed mean. There are certainly societies (the UK isn't the worst example, but it does exhibit some of the traits) where an outrageous amount of wealth is found in the top 1% and this can most definitely skew an arithmetic mean. This is a frequent mistake in stats when people talk about "the average" 99 times out of 100 (couldn't resist that), they're actually talking about the arithmetic mean, which is just one measure of an average (you all know that) but it's amazing how many times we allow ourselves to get sucked into this error. You might even find the mode is the better indicator in this case Silly example - but it's often used to illustrate this kind of thing: How many arms has the average person got? If you used the mean the answer would be 1.999 If you used the mode it would be 2 If you were manufacturer of shirts and decided to make products based around the average (as defined by the mean) would that be sensible? [/QUOTE]

Verification: 5 + 3 = ?