Average salary

[MrRobin]

The median and the range are better at describing skewed distributions than the mean.

Can we just leave it at that?

[scooterscot]

In Berlin we managed to wangle a suite in the Hotel Berlin for 35 Euro a night - bloody breakfast cost more but we wangled that somehow - current Ms stek speaks fluent German so she did her stuff - 7 Euro each ISTR....

If I had a contract in Berlin I'd quite happily like my life out in that hotel....

Berlin is my fave city, dunno why - just is!

Berlin is a great place to live, if you can get a job there.

Good point to rant about, cost of hotels compared to the UK. A night in a hotel near Bedford cost me £115 and it was an utter hole, I seriously would not allow dog to stay there, but somehow for some strange weird reason the hotel feels they are perfectly just charging that. It's a huge crock.

Rant number II, travel lodge must be sitting on the largest wod of cash ever. They're buying up every empty space in Edinburgh and turning it into a hotel even if it's a 3 room joint you lazy council $@£$ get off you're £$%$ and stop this madness before we're hip deep in chavs

[ThomasSoerensen]

The median and the range are better at describing skewed distributions than the mean.

Can we just leave it at that?

and get back on topic?

what a preposterous suggestion

[sasguru]

Surely you are talking about yourself. As what Doggy Styles wrote is correct.
You are a numpty

Doggy Styles says that the median is not really an average.
And I'm the numpty?

[DimPrawn]

Doggy Styles says that the median is not really an average.
And I'm the numpty?

Is it someone who speaks to dead people?

[Doggy Styles]

Another statistical illiterate.

I can't see what was wrong with anything I said there.

I even conceded that my example was in extremis and only presented to better illustrate the point, before applying the point to real values.

You can find decriptions of average, median and mode online.

[RichardCranium]

Median is a type of Average you muppet.

Not really.

There are a number of different ways of specifying an average because they are needed in different contexts.

The mean average is crap and meaningless for specifying salaries when done across the whole country. It is meaningful as the average for project managers or bus drivers.

The mean average gets used for all sorts of inappropriate uses. E.g. car parking space width.

The modal average is great for determining the stock in a stack-em-high-sell-em-cheap outlet: it's the "what do most customers want and bugger the rest" philosophy.

The median is useful because it does not get skewed by extremes. The median household income will not change if the top 100 paid people's pay goes up. Nor will it change if the minimum wage goes up. It effectively eliminates "outliers". So if you want the average height of boys in a class of 15 year olds, using the median will eliminate the two baseball-playing freaks in 5C.

Knowing families have 2.4 children when selling "Family Tickets" is no use - knowing the median family has 2 children tells you at least half of all families will be able to buy a "Family Railcard". You cannot specify "Valid for up to 2.4 children" in the T&Cs!

The median tells you the mid-point that is, half the population is below and half is above. Using it as a barrier is "fair" or equitable. It also gives a real world value: the mean number of legs for the UK population is about 1.98. The median number of legs is 2. (However, the mean number of fingers is probably over 10, the median number of fingers probably below 10 - the modal average will be 10.)

There is no point ordering helmets for the Army in mean hat size 7.31456; the median hat size 7½ would be cheaper.

The median IS an average and it has its uses.

[sasguru]

Quite right RC - and this is more than merely a pedantic discussion.
Consider for example that I read (in the Economist I think it was) that the median income in the US is half the mean.
I can't vouch for the veracity of this fact but it shows that the type of statistic you use can lead to very different results.

[ThomasSoerensen]

Quite right RC - and this is more than merely a pedantic discussion.
Consider for example that I read (in the Economist I think it was) that the median income in the US is half the mean.
I can't vouch for the veracity of this fact but it shows that the type of statistic you use can lead to very different results.

which is exactly why I specified in my original post that one number was median and the other number was mean. So people could keep the peculiarities in mind when discussing the numbers, which are acturally the interesting part of my original post.

[Bagpuss]

Not really.

If two people earned £10 a year and one earned £300,000:

• Their average wage is a little over £100,000 per annum
• Their median wage is exactly £10 per annum

That's obviously an extreme case, but it illustrates how the millionaires and billionaires of the country push the average wage higher than the median.

The mode would be an interesting one. I'd guess it could be the minimum wage.

Alf is exactly right, the median is a type of average, the other two being the mean and the mode. Numerically illiterate people e.g. Journalists usually incorrectly call the mean 'the' average whereas it could be any of the three.


PS mean of those numbers isn't a little over 100k nor have you got the median correct.