In the textbook All of Statistics: A Concise Course in Statistical Inference by Larry Wasserman, the definition of minimal sufficient is given as follows:

9.35 Definition. A statistic $T$ is minimal sufficient if (i) it is sufficient; and (ii) it is a function of every other sufficient statistic.

It then gives the following theorem:

9.36 Theorem. $T$ is minimal sufficient if the following is true: $$T(x^n) = T(y^n) \ \text{if and only if} \ x^n \leftrightarrow y^n.$$

Why does this theorem have these exponents of $n$? Having these exponents of $n$ doesn't really make sense to me.


In that book, $n$ random variables $X_1, \dots, X_n$, constituting the "data set", are represented as $X^n$. Whereas realisations of those $n$ random variables $x_1, x_2, \dots , x_n$ are represented as $x^n$.

There is no exponentiation occurring.

  • $\begingroup$ Ahh, that makes sense. I'm jumping through the textbook to parts of interest, rather than reading from the beginning, so I missed that. $\endgroup$ Apr 20 '21 at 1:05
  • 1
    $\begingroup$ That book can be quite terse, but appropriate if you require stats knowledge only as a pre-requisite for machine learning work. He has an accompanying publically available Youtube course here: youtube.com/playlist?list=PLJPW8OTey_OZk6K_9QLpguoPg_Ip3GkW_ And lecture notes on his CMU course page. $\endgroup$
    – microhaus
    Apr 20 '21 at 1:12

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.