Halifax -1.9% MoM, -17.5% YoY -21.3% from peak.

Really2

Sir_Humphrey wrote: »

For a nationwide picture, 1000-1500 is a large enough sample to get an accurate measure. You are incorrect about opinion polling, if those cities were Glasgow and Manchester you would not get an accurate picture, but a sample of 1000-1500 is enough to smooth out those issues (to about 95% confidence).

Sorry you are talking out your !SS there.

You could judge the average price drop of every house in england by selecting 1 in 25,000?

Not taking in to account where it is what type or the size?

What is the most accurate opinion poll the 1 with 1000 people or the general election?

House price is not opinion it is based on financial facts which very from region to region and many other factors.

You are obviously drilled that opinion poles are always right from your public sector back ground.

But I can tell you in financial terms you wrong to think limited data presents an accurate figure.

Believe me I write 1/4 financial reports to predict the next 1/4 year etc.

You can't use a small number of samples, are you really that naive, I never thought you were.

Could you predict the correct number I am thinking from 1-25,000

stephen163

neas wrote: »

I have to laugh at SKY news broadcasting:



Riiiiight... so they fall at almost 2% (higher than the average drop month on month)... and the logic is its showing signs of stopping? Wouldny signs of stopping be a 0.9% then a 0.5% then a 0.2% then a 0.1% fall etc etc?

Not just a constant fall. Seriously these guys are full of doo doo..

Using MoM data in that way is folly anyway! They should be shot for poor journaism

Sir_Humphrey

Really2 wrote: »

Sorry you are talking out your !SS there.

You could judge the average price drop of every house in england by selecting 1 in 25,000?

Not taking in to account where it is what type or the size?

What is the most accurate opinion poll the 1 with 1000 people or the general election?

House price is not opinion it is based on financial facts which very from region to region and many other factors.

You are obviously drilled that opinion poles are always right from your public sector back ground.

But I can tell you in financial terms you wrong to think limited data presents an accurate figure.

Believe me I write 1/4 financial reports to predict the next 1/4 year etc.

You can't use a small number of samples, are you really that naive, I never thought you were.

Could you predict the correct number I am thinking from 1-25,000

Stephen163 put it a lot better than me. Selecting a 1000-1500 (probably smaller than Halifax's sample) is not the same as picking 1. It is the size of the sample, not the sample proportional to population that matters. The mix adjustment and the suchlike weights the sample to a similar end to the stratified random sampling used in opinion polls.

Of course general elections are more accurate than opinion polls as it uses the whole population who actually votes, but it would far too expensive to interview everyone for an ICM poll. This is not a public sector point of view, but an understanding stats POV.

Really2

stephen163 wrote: »

A sample size of 1000 is too small but they compensate for this by stratifying and weighting. Statistically, it is true that the more samples you get the more accurate your data will be, however, once you pass a certain sample size, the gains get smaller and smaller so in practice, a sample size of a small fraction of your population is all that is required to satisfy a certain confidence interval.

But being 5-10% out is not good not in finance.

1000 woul take what a sample 200 of each different type of property (not even taking size in to consideration)

The only true average house price in the UK would exisit if every house was sold at the same time. Wighting is guessing as this is not oppinion.

motch

i'll have to go with really2 on this one.

38000 you'll get a reasonably accurate House price Data.

1000-1500 you definetely wouldn't, due to lots of areas only selling 1 or 2 properties. just wouldn't be accurate.

Sir_Humphrey

motch wrote: »

i'll have to go with really2 on this one.

38000 you'll get a reasonably accurate House price Data.

1000-1500 you definetely wouldn't, due to lots of areas only selling 1 or 2 properties. just wouldn't be accurate.

You expect it to be the case, but lots of very clever maths professors over many decades have shown otherwise. The extra accuracy would be very small indeed.

That it is not good enough for financial trading just proves that financial trading is gambling rather than stats are invalid.

Really2

Sir_Humphrey wrote: »

Stephen163 put it a lot better than me. Selecting a 1000-1500 (probably smaller than Halifax's sample) is not the same as picking 1. It is the size of the sample, not the sample proportional to population that matters. The mix adjustment and the suchlike weights the sample to a similar end to the stratified random sampling used in opinion polls.

Ok then I will let you pick 1000 numbers from 25,000,000 to find the 1000 I am thinking of.;)

Lets put it this way to you guys.

Do you think every house went up 1.9% earlier this year?:rolleyes:
Does a year give a better idea than a month?:rolleyes:

Low samples do not give an acurate idea of the market what so ever that why the odd spike or massive drops are happening.

100,000 houses sold gives a lot better indication than 1,000 if you belive otherwise I worry for the country.

Sir_Humphrey

Really2 wrote: »

Ok then I will let you pick 1000 numbers from 25,000,000 to find the 1000 I am thinking of.;)

Lets put it this way to you guys.

Do you think every house went up 1.9% earlier this year?:rolleyes:
Does a year give a better idea than a month?:rolleyes:

Low samples do not give an acurate idea of the market what so ever that why the odd spike or massive drops are happening.

100,000 houses sold gives a lot better indication than 1,000 if you belive otherwise I worry for the country.

They could have done, but the probability would probably be less than 0.01%. On the other hand they could have gone down by more than 1.9%. I do not know the exact margin of error.

If 10000 were the population figure for houses, the you would be correct. You are correct that 10000 would be more accurate. The point is that the degree of extra accuracy would be low. In fact if Lloyds/HBOS have 30% of the market, and there are about 35k approvals, the actual sample is probably about 10000, not 1000.

There is large margin of error with MoM stats, which is why as I have repeated said in the past not to get too worked up about them.

Kenny4315

What clear is, nothing is clear .... :rolleyes:

Looks like the 35% down from peak is a go-er again. :money:

I'll just get my trolley polished and ready for action .;)

Really2

Sir_Humphrey wrote: »

That it is not good enough for financial trading just proves that financial trading is gambling rather than stats are invalid.

What the hell are you on about financial companies use 100% stats not samples.

To work out a companies credit rating do you look at that company actual accounts or 1000 UK companies and average it out.

SH admit you are a bit lost on this one.