6
$\begingroup$

"The number of degrees of freedom is a measure of how certain we are that our sample population is representative of the entire population - the more degrees of freedom, usually the more certain we can be that we have accurately sampled the entire population."

Can someone please explain this statement?

Improve this question
$\endgroup$
6
  • 5
    $\begingroup$ It is essentially nonsense. You can have an enormous sample that is highly unrepresentative. You can have a very small sample that is representative. Also, the idea of representative sample is not at all clear-cut. See the papers by Kruskal and Mosteller cited in stats.stackexchange.com/questions/16847/… for more on "representative sample". In short, size of sample and number of degrees of freedom associated with a particular procedure are quite unconnected with representativeness, however the latter is defined. $\endgroup$
    – Nick Cox
    Nov 15, 2014 at 15:08
  • $\begingroup$ @Nick Cox What is then purpose of the degrees of freedom? $\endgroup$
    – Quirik
    Nov 15, 2014 at 15:11
  • 1
    $\begingroup$ Many posts here on degrees of freedom. Please use the resources already available. $\endgroup$
    – Nick Cox
    Nov 15, 2014 at 15:17
  • 1
    $\begingroup$ Where does the quote come from? It doesn't make any sense to me. $\endgroup$
    – Glen_b
    Nov 15, 2014 at 15:19
  • $\begingroup$ chem.utoronto.ca/coursenotes/analsci/StatsTutorial/DegFree.html $\endgroup$
    – Quirik
    Nov 15, 2014 at 15:20

1 Answer 1

Reset to default
13
$\begingroup$

There is a sentence prior to the passage quoted by the OP that I believe helps to interpret this:

In statistics, the number of degrees of freedom (d.o.f.) is the number of independent pieces of data being used to make a calculation. (...). The number of degrees of freedom is a measure of how certain we are that our sample population is representative of the entire population - the more degrees of freedom, usually the more certain we can be that we have accurately sampled the entire population.

So here

"more degrees of freedom"$\equiv$ "greater number of independent pieces of data"

This starts to sound familiar, since it points to the size of a sample of independent draws from the population. Moreover, on focus here are experimental data, so all nice properties I guess are assumed to be guaranteed, and therefore the larger the sample size of independent pieces of data, the more strongly the consistency property of estimator will actually emerge and reflect upon the estimates obtained.

So it appears that, for the author of the passage, the logical chain here goes as follows:

"more degrees of freedom"$\equiv$ "greater number of independent pieces of data"

and

"greater number of independent pieces of data" $\Rightarrow$"greater accuracy in recovering the population moments from the data"

and

"greater accuracy in recovering the populations moments from the data" $\Rightarrow$ "the more representative of the population is the sample"

So

"more degrees of freedom"$\Rightarrow$ "the more representative of the population is the sample"

It appears therefore that the author of the passage employs the term "representative sample" with the meaning of "miniature population" or of "typical or ideal case", to follow the typology of Kruskal and Mosteller as re-relayed here.

Improve this answer
$\endgroup$
2
  • $\begingroup$ What does it mean " greater accuracy in recovering the population moments from the data"? $\endgroup$
    – Quirik
    Nov 15, 2014 at 17:44
  • $\begingroup$ @Navi It means "the estimators are "asymptotically consistent" and so as the sample size increases a) it is more and more probable that the obtained estimate will be closer and closer to the true value and b) that the variance of the estimator will be lower (in most cases). $\endgroup$
    – Alecos Papadopoulos
    Nov 15, 2014 at 18:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.