Tagged Questions

Bias, in a statistical framework, means that an estimate of a parameter has an expected value that is not equal to the actual parameter value. There is often a tradeoff between bias and variance - low variance estimators may be more biased than high variance ones.

54 views

Proper Sampling - can I collect a two-group sample this way without issues?

I need to collect a two group sample for a comparison analysis (perhaps using logistic regression). The population that I need to extract a sample from is all firms from country A with activities in ...
56 views

Detrending Discrete Data

I am trying to detrend some discrete data and I am having difficulty finding a model to describe the trend. There is a number of discrete data points and there is a linear error being introduced with ...
28 views

Data transform introducing a bias in kriged geostatistical model?

I am kriging a 3D geospatial model of saturation data. There is water saturation ($SWT$), gas saturation ($SGT$) and oil saturation ($SOT$). A constraint is that the saturations must add up to one. ...
56 views

Estimator bias without a closed form?

Given a regression loss function $l(Z,\beta)=||Y-Z\beta||_2 + \lambda \beta^TD\beta + r(X,Z)$ where $X$ is the predictor matrix, I would like to estimate a $Z$ that minimizes the above loss in a ...
61 views

The bias of Zellner estimators in dynamic SUR models

I have been playing around with a seemingly unrelated regression (SUR) estimation. However, for dynamic SUR models it is known that -- analogous to the ARIMA case -- an OLS/GLS estimate is biased. For ...
78 views

Is it OK to use the CLT to create a normal distribution where there is none?

I have some data that looks like this: Procrastinator has come up with one good suggestion for how to test hypotheses under this distribution, but it relies on some guesswork to fit constants. I ...
111 views

Generalized Linear Models and Curse of Dimensionality

I was wondering what happens to bias and variance of GLM estimates as dimensionality approaches the number of training data points? Specifically in Linear Regression and Poisson Regression? I know ...
20 views

How to calculate differential item functioning when factor structure differs between groups?

I want to test for differential item functioning on a self-report measure between two groups (i.e., one with a disease one without). Differential item functioning refers to a measurement bias wherein ...
29 views

Correcting for Bias due to Larger Populations

I'm mining social networks point data to use with GIS (Geographic information system). Obviously there are going to be more posts in areas with higher population which, if uncorrected, would end up ...
26 views

Identifying potential for missingness in a seasonal time series to bias period-averages

I have high-frequency time series (observations every few minutes) for which I wish to compute daily averages. The data exhibit a strong diel cycle. Sometimes observations are missing in the time ...
43 views

R package random forest: can corr.bias be harmful?

There exists an option corr.bias for making regression analyses with the R package randomForest. The manual warns “Experimental. Use at your own risk.” What does corr.bias precisely do and why can it ...
53 views

Unbiased estimate of the semi-partial correlation

Is the sample semi-partial correlation a biased estimate of the population semi-partial correlation? If it is biased, what is an unbiased estimator of the population semi-partial correlation? Are ...
101 views

“Rule of Thumb” method to adjust for overstatement bias on a likert scale

I have a relatively novice question which I have been attempting to peruse the literature for the past hour to address without much success. I am attempting to test a product concept in a survey for ...
89 views

Given $y=f(x)+\epsilon$ where $x=(x_1,\dots,x_p)$, $f$ is highly non-linear and two different estimators: $\hat{y}=\hat{M}(x)$ $\hat{y}=\hat M_1(x)+\hat M_2(x)$ where $M_1$ is a simple (biased) ...
31 views

OLS unbiased for all sample sizes

I am pretty positive that OLS regression produces unbiased estimates for all sample sizes, even though the variance about those estimates might become very large when sample sizes become small (e.g. ...
39 views

Does Bayesian data analysis take into account estimates' bias?

For example, the standard deviation is known to have bias that depends on the number of samples observed. If I wanted to do Bayesian inference on the SD of samples from two populations, and have ...
28 views

Bias-variance tradeoff — bias or variance effect

I understand that supervised learning is associated with an error that can be split between bias and variance: $MSE = b^2 + var$ What does bias and variance account for intuitively ? I ...
35 views

Interpretation of quantile regression when high quantile estimates are lower

Suppose I estimate a multivariable quantile regression $Q_Y(\tau | X) = \alpha(\tau) + \beta(\tau)X + \epsilon(\tau)$. Note that $X$ is a vector of independent variables. Suppose I then 'plug in' my ...
34 views

RMSE when parameter is observed with uncertainty

Here is my situation: I want to know the value of some parameter $\theta$ and I have a bunch of different estimates of it. Because each estimate is drawn from a different sample, I observe the ...
89 views

Measurement bias with Likert items and dichotomous items

I'm hoping someone here might be able to answer a question I am struggling with. I'm trying to find the best way to test differences across and within 3 conditions, where I am manipulating my ...
45 views

Detection bias and ascertainment bias

Are detection bias and ascertainment bias the same concept? I saw that allocation concealment is to reduce detection bias in some website, and to reduce ascertainment bias in some other website. ...
29 views

Panel data by subgroups

Panel data by subgroups I have a panel dataset on 200 firms showing the turnover of financial managers and firm performance for ten years. Some of the individuals are fired and some get promotion ...
45 views

Bias of maximum likelihood estimators for logistic regression

I would like to understand a couple of fact on maximum likelihood estimators (MLEs) for logistic regressions. Is it true that, in general, the MLE for logistic regression is biased? I would say ...
199 views

Question about predictive bias - intercept and slope bias

I am slightly confused on how to determine a slope and intercept bias. I have an assignment where i am supposed to conduct a gender predictive validity bias analysis. However, my lab handout and the ...
49 views

Consequences of non-stationarity on panel regression estimates

What are the consequences of including a non-stationary variable on a panel regression's slope estimates and their standard error estimates? I am thinking of both Pooled OLS and Entity Fixed Effects. ...
58 views

Unbiased estimator of weighted sum of two poisson variables

Suppose that $X_1$ and $X_2$ are two random variables sampled from a Poisson distribution with parameter $\mu$. Let $T_1=\bar{X}$ be the sample mean and let $T_2=(1/3)X_1 +(2/3)X_2$. Are T1 and T2 ...
80 views

How do I explain that software implemented model selection procedures should not be used unsupervised?

I know that people generally say that procedures which select a model based on information criterion lead to inconsistent model selections. I read a paper by Leeb and Potscher (2005), MODEL SELECTION ...
140 views

What are the conditions where we can regress non-stationary variables?

Obviously there are certain spots where it's okay to include a non-stationary predictor variable in a linear regression model. For example, a dummy variable interacted with a stationary variable must ...
117 views

Non-response bias

I have the following question: In a survey, a simple random sample of 1,000 households was drawn to determine the distribution of household size in a city. Interviewers were required to visit ...
Biased variation of $\chi^2$ statistic?
I've found a variation of the $\chi^2$ statistic that looks like this: $\chi^2 = \sum\limits_{i=1}^N\,\chi_i^2 = \sum\limits_{i=1}^N\,\frac{(\log m_{i}- \log n_{i})^2}{\sigma_{i}^2}$ where ...