Questions tagged [inference]
Drawing conclusions about population parameters from sample data. See https://en.wikipedia.org/wiki/Inference and https://en.wikipedia.org/wiki/Statistical_inference
3,176 questions
1
vote
0
answers
45
views
Conceptually Interesting Applications of the Delta Method
What are some nice, insightful applications of the delta method?
From Casella & Berger's Statistical Inference book (2nd ed.), the following example appears under The Delta Method:
Example 5.5.19 ...
- 11
1
vote
1
answer
35
views
Survey sample size required to compare proportions
I have survey data across two years (random selection of participants), one of the questions determining the income group of repondents (we distinguish 4 groups). Across years we might observe ...
- 191
0
votes
0
answers
23
views
Difference between Inference, Decision, Estimation, and Learning/Fitting in Generalized Decision Theory?
I am trying to strictly define the relationships between Inference, Decision, Estimation, and Learning/Fitting using the framework of Generalized Bayesian Decision Theory (as taught in MIT 6.437).
Set-...
- 7,398
1
vote
0
answers
28
views
Multiple Treatments on Multiple Units
The main question is: How to handle causal inference with multiple units experiencing multiple (non-permanent) treatments over time?
I am working on a causal inference problem where I have multiple ...
- 172
14
votes
6
answers
948
views
Why does classical testing privilege Type I error (vs. the symmetric Bayesian view), why is status quo given privilege in traditional decision theory?
I'm trying to reconcile the classical Neyman-Pearson (NP) approach to hypothesis testing with the more symmetric, decision-theoretic view we use in my 6.437 (Inference & Information) class.
In our ...
- 7,398
2
votes
1
answer
64
views
Why do we use convergence in probability to define consistency of an estimator?
I'm trying to find an explanation for the reason we define consistency of an estimator as a probability convergence. Basically I'm interest in understanding why we don't use almost sure convergence as ...
- 21
0
votes
0
answers
20
views
Does a fully deterministic Pocock & Simon minimization affect variance estimation and inference validity (ignoring selection bias)?
I have a question about covariate-adaptive allocation in clinical trials.
Suppose we use a Pocock and Simon minimization procedure without any random component: that is, a fully deterministic ...
1
vote
0
answers
12
views
Graph-RAG-CV (Fusion Reranker No-Reinfer) [closed]
Short
Can I build a Graph-RAG pipeline that applies a reranker and top-K after vector search without re-running the full inference loop? I also need to continuously ingest VLM/CV monitoring logs and ...
1
vote
1
answer
102
views
Maximum likelihood estimation with the likelihood function as a function of random variables
I have seen many texts use this notation where they define the likelihood function as a function of random variables and the unknown parameter, rather than a function of the data and the unknown ...
- 115
1
vote
0
answers
34
views
$X_1,...,X_m$ follows iid $\text{Bin}(n,p)$, $0<p<1$ . $T=\sum X_i $. What is the UMVUE of $q/p$ where $q=1-p$? [duplicate]
I know that $T \sim \text{Bin}(mn,p)$ and it's also complete and sufficient. For a UMVUE I need a function of $T$. But I'm bothered by $1/p$ situation. Would $mn/T$ be an UE of $1/p$ just as $T/mn$ is ...
- 21
0
votes
0
answers
35
views
What are the limitations of AUC/ROC when used to evaluate effects over time (e.g., drug, electrophysiological signals) in inferential models?
I’ve been reading some papers in the neuroscience field, and I don’t quite understand the widespread use of AUC/ROC to test for group differences when analyzing neuronal firing over a range of seconds ...
- 1
0
votes
0
answers
58
views
What does it mean for a prior to be "reliable" for the parameter?
In reading Lehmann & Romano (1998)(which I assume has not changed much in the 2022 edition), they mention, in passing (I believe in Chapter 3, but I cannot find the exact spot), that prior ...
- 4,264
7
votes
1
answer
167
views
Statistical modeling with only a single data point
The vast majority of statistical literature involves having a dataset which can be partitioned into $n$ data points, $\mathbf{x} = \{x_1,...,x_n\}$ constructing a model for the individual data ...
- 121
0
votes
0
answers
32
views
Is an explicit "treatment" variable a necessary condition for instrumental variable analysis?
I'm trying to model the causal impact of our marketing efforts on our ads business, and I'm considering an Instrumental Variable (IV) framework. I'd appreciate a sanity check on my approach and any ...
- 65
1
vote
1
answer
101
views
How do I evaluate the covariance in the frequentist paradigm?
I am doing inferencing with a frequentest paradigm. I don't think it matters, but the context of this is to understand information geometry.
I would like to understand the Frequentist covariance ...
- 111