Put , and . Apart from a data-only constant, the log-likelihood is
The natural parameters in the ambient two-parameter exponential family of normal distributions are and , restricted to the nonaffine curve , . Its parametrization has rank one. Thus this is a one-dimensional curved exponential family inside a two-dimensional full family.
The Fisher-Neyman factorization theorem makes sufficient. For two samples , the parameter-dependent part of their log density ratio is . A polynomial in is constant on only if both nonconstant coefficients vanish. The likelihood-ratio criterion for minimal sufficiency therefore proves is minimal sufficient.
Write with independent . Since , has a probability distribution independent of : it is an ancillary statistic. The denominator is nonzero almost surely. The Cauchy-Schwarz inequality gives ; on the requested positive branch, and .
The likelihood equation is . Its positive root is
There is exactly one positive root. The sign of the score function changes from positive to negative there, and at both ends of when , so it is the unique maximum-likelihood estimator. The useful identities are
Substitution yields, dropping the constant ,
Differentiating twice and using at the fitted value gives the observed information
The scalar P-star approximation uses . Consequently, with its normalizing factor explicitly retained,
Here is chosen to make the density integrate to one, and is independent of . To see this without assuming an approximate normalizer, define
The integral is finite and positive, and normalized is exactly
Indeed .
There is a direct check of this conditional scale density in a curved Gaussian family. In polar coordinates , the sample density times volume element is proportional to . Conditioning on fixes . The remaining angular factor is independent of , while . Thus the exact conditional radial density has the same normalized form. For the positive ancillary branch fixes the positive sign and the same one-dimensional radial calculation applies. In this scale model the normalized expression equals the exact conditional density, not merely its leading approximation.
For , compute its conditional distribution by one-dimensional quadrature:
It is zero for ; with the conditional-density calculation just given, this normalized formula is exact for almost every admissible . Numerically, subtract the maximum log integrand before exponentiating to avoid overflow. The sampling probability still depends on the stipulated true ; replacing it by a fitted value would be a separate plug-in inference step.
The joint likelihood is , so the factorization criterion for sufficiency proves sufficiency. For any two positive sample vectors its likelihood ratio is , independent of exactly when their sums agree. The likelihood-ratio criterion for minimal sufficiency therefore makes minimal sufficient.
Its gamma distribution has density on . If is integrable for every and has expectation zero for every , then
By uniqueness of the Laplace transform, almost everywhere. One may first multiply by to obtain an integrable function and use transform parameters ; this handles the transform's integrability hypothesis. Since the gamma densities are positive on , almost surely for every . is complete as well as minimal sufficient.
A complete statistic for a family if every measurable with and for all satisfies almost surely under every . It is boundedly complete if the same implication is required only for bounded measurable . Completeness therefore implies bounded completeness.
The vector is multivariate normal with mean zero and covariance
Its determinant is and its inverse is . Hence the joint density is
The Fisher-Neyman factorization theorem gives sufficiency. For minimality use the likelihood-ratio criterion for minimal sufficiency. The density ratio for is independent of precisely when
is constant. Multiplying by shows that its constant would produce a quadratic term on the right whereas the left is linear, so that constant is zero. The remaining coefficients give . Thus is minimal sufficient statistic.
Each is standard normal, so . The integrable, nonzero function has zero expectation for every parameter, proving that is not complete.
By sufficiency, choose a version common to all parameters. The tower property gives
On the event we necessarily have , so there almost surely. This event has positive probability under every parameter because the joint normal density is strictly positive on a ball about the origin. Thus is not identically . The bounded function has zero expectation for every parameter but is not almost surely zero. Consequently
Write , and . The joint probability density factors as
By the Fisher-Neyman factorization theorem, is a sufficient statistic. The same factorization shows that the smaller statistic is also a sufficient statistic. This gives the sufficient statistic for a symmetric uniform sample.
In fact, the likelihood-ratio criterion for minimal sufficiency identifies as a minimal sufficient statistic: two sample points with the same have identical likelihoods for all ; with different , their likelihoods have different parameter supports, and so are not proportional as functions of . Equivalently, the condition that for every and some positive constant holds exactly when their absolute maxima agree.
To see rigorously that cannot be recovered from almost surely, fix any . The sample has a distribution invariant under simultaneous sign reversal. If almost surely, the same identity would hold for the sign-reversed sample on another probability-one set. Since is unchanged by sign reversal, it would follow that
and hence almost surely. But this event has probability zero for a finite sample from a continuous uniform distribution: it entails for some pair of indices, including the possible case . Therefore is not a function of the sufficient statistic . The pair of extrema is not minimally sufficient, for every , even though it is sufficient.
A sufficient statistic has a conditional distribution of the full data given its value that is independent of the unknown parameter. A minimal sufficient statistic is sufficient and is a function of every other sufficient statistic, up to null sets. For a dominated family with common positive support, the likelihood-ratio criterion for minimal sufficiency states that is minimal sufficient if holds exactly when is independent of .
For , the joint probability mass function is
The Fisher-Neyman factorization theorem proves sufficiency of . For two samples, the likelihood ratio is , independent of exactly when their sums agree. Thus is minimal sufficient. Any one-to-one transformation of it is also minimal sufficient, so choose
This chosen estimator is an unbiased estimator, since each observation has mean . If instead the statistic is reported as , its expectation is , and as an estimator of it has bias (zero only when ). Being a minimal sufficient statistic is unchanged by this rescaling; being an unbiased estimator is not.
For , define the class degree sum
and define the number of within-college friendships by
Each between-college edge contributes once to the degree sum of each endpoint's college, while each within-college edge contributes twice to its college's degree sum. The model-3 likelihood can therefore be written
The Fisher-Neyman factorization theorem shows that is sufficient.
For two networks , the likelihood ratio is
Because the parameters range independently over , this ratio is constant in exactly when every displayed difference vanishes. The likelihood-ratio criterion for minimal sufficiency therefore gives the answer
This is the degree-sum sufficient statistic for an additive logistic network model.
Expanding the exponent gives
The Fisher-Neyman factorization theorem therefore shows that is sufficient.
It is also minimal sufficient. For two sample points ,
where is independent of . This ratio is independent of exactly when . The likelihood-ratio criterion for minimal sufficiency applies, proving the claim and the general normal sample sum with known variance result in this case.
A sufficient statistic is minimal sufficient if it is a function of every other sufficient statistic: whenever is sufficient, there is a function such that almost surely, up to the usual null-set qualification. Thus its level sets give the coarsest sufficient partition of the sample space. In a dominated family with positive densities, the likelihood-ratio criterion for minimal sufficiency says that is minimal sufficient when
For , the density is
so the Fisher-Neyman factorization theorem proves that is sufficient.
Moreover,
is independent of exactly when , equivalently . The likelihood-ratio criterion for minimal sufficiency therefore proves that