any statistics genius available?

[historian]

Quick question if you can answer. The govt stats excludes those whose unemployment benefits run out; those who give up looking or those who have never worked living on subsistence levels (growing food etc...such as in the appalachians) and does not include former highly paid professionals now earning far less at jobs for which they are overqualified.
Is it fair to claim that the results are therefore wrong (from a statistical point of view?)

[Jackpine Radical]

1. Short answer: No.

Statistics can't answer that question since it's not really an empirical question, and hinges on what you mean by "wrong." People can, however, draw wrong conclusions from any set of data because they misunderstand the data.

[historian]

3. by wrong i mean

it excludes a certain part of the population so it is therefore not correct. Could it be called manipulation of numbers then?

[Eye and Monkey]

7. I think you're on the right track.

Get out your stats text, look in the index under "error" and "bias".(or google/yahoo).

[Eye and Monkey]

5. And the longer answer: you might change the question.

In the instance you've sketched above, it seem s that the problem is not one of statistical analysis. The problem is more along the lines of defining what numbers (or subpopulations) will be included in the analysis - and the implications that users of the analysis purports it indicates.

[TruthIsAll]

2. Of course they are wrong. 5.7% unemployed is a total crock.

The 5.7% are those who are receiving unemployment benefits. You must add those whose benefits have run out; those who have had to accept p/t clerical, fast food, maintanence work, etc. at minimum wage; those who have stopped looking because their jobs have been outsourced forever.

So now we are up to 10-12%. Maybe more.

[historian]

4. i understand that

but i have to present this to a professor of stats

[Jackpine Radical]

8. It's a matter of definitions, particularly of operational definitions,

and of the universe to which you are trying to generalize. If the question you set out to answer has to do with unemployment as defined by those who are seeking work through the unemployment office, then your measure is accurate. If you are trying to talk about "unemployment" in some global sense that includes discouraged workers, underemployed people, etc., then the measurement method is obviously not appropriate to the question. The problem is in the mismatch between the method and the question. Neither is right or wrong in itself, but the combination is wrong.

I have no idea if that helps or not.

[dsc]

6. clearly the statistics are wrong

but the question that I think you are asking is "Are they more wrong now than they were before." Where I would concentrate effort to answer that question is on the refusual to extend unemployment benefits. This is the first recession where they weren't extended without a fight and where they still aren't extended again now. I hope that helps. The other factors would likely be the same throughout history.