Parameters, Estimates I lack some knowledge in the concepts of parameters, estimates and moment (math and stats). I can't find an online easy-to-understand source of information about these concepts. Would you help me with a link or even better with a speech ?
Here is one thing that I'd aimed to understand:
Source: a book called Principles in Population Genetics

$\sigma_x^2 = \text{E}[x-\text{E}(x)^2]$  
$\text{Var}(x) = \text{mean}(x^2) - \text{mean}(x)^2$  
But it is a biased estimator because...
$\text{E}[\text{Var}(x)] = (n-1)\sigma^2/n$
A non-biased estimator for $\sigma^2$ is therefore
$[n/(n-1)]\text{Var}(x)$

What is the meaning of all this ?
 A: This really starts to look like a "takes a book to answer" question, but I am answering anyway because this is too long for a comment
It's difficult to tell the level of detail you require.
The notation is bad. This is a little nearer to what it should look like, if more standard notation was used:

Let $X$ be a random variable with some distribution, and $\boldsymbol{x} = (x_1, ... , x_n)'$ be a set of independent observations each distributed as $X$. Let an overline (bar) over something represent its sample mean, so $\bar{\boldsymbol{x}}$ is the sample mean of $\boldsymbol{x}$. Define
$\sigma_x^2 = \text{E}[\boldsymbol{x}-\text{E}(\boldsymbol{x})^2]$  and we might estimate this from the sample by
$\widehat{Var}(\boldsymbol{x}) = \overline{(\boldsymbol{x}- \bar{\boldsymbol{x}})^2}  = \overline{\boldsymbol{x}^2} - \bar{\boldsymbol{x}}^2$  
But it is a biased estimator because...
$\text{E}[\widehat{Var}(\boldsymbol{x})] = (n-1)\sigma^2/n$
A non-biased estimator for $\sigma^2$ is therefore
$\frac{n}{n-1}\widehat{Var}(\boldsymbol{x})$

But if it were me, I would probably go about introducing this stuff quite differently.
Also, it would be more usual still to write $s^2_n$ (or perhaps $\hat\sigma^2_n$) for that first (biased) sample variance estimate and  $s^2_{n-1}$ for the second (or $\hat\sigma^2_{n-1}$)).
As for the definitions of things like variance and bias, they're probably better looked up:
http://en.wikipedia.org/wiki/Bias_of_an_estimator
http://en.wikipedia.org/wiki/Variance
and so on.
Your best bet is to find two or three books that you like on basic mathematical statistics and work through them. There are many dozens, probably hundreds (Wackerly, Mendenhall  and Scheaffer; Devore; Peter Smith; Mood, Graybill and Boes; Hogg and Craig; Kalbfleisch etc etc)
