Though not applicable in every field, occasionally statisticians run into problems with small sample sizes. Experiments with small sample size tend to be more unstable and biases, resulting in poor parameter estimations. These same concerns apply to experimental designs. This paper investigates the "properties of designs for small experiments" and focuses on minimizing the mean squared error in small experiments.
Thursday, March 1, 2012
Source (3/2)
Russell, K.G., Eccelston, J.A., Lewis, S.M. and Woods, D.C. "Design considerations for small experiments and simple logistic regression." Journal of Statistical Computation and Simulation. 79.1 (2009): 81-91. Electronic.
Though not applicable in every field, occasionally statisticians run into problems with small sample sizes. Experiments with small sample size tend to be more unstable and biases, resulting in poor parameter estimations. These same concerns apply to experimental designs. This paper investigates the "properties of designs for small experiments" and focuses on minimizing the mean squared error in small experiments.
Bias is a concern in statistics. When I work on the theory behind statistical design, it is an important factor to investigate. This paper provides a good mathematical approach to finding optimal designs for experiment while controlling for bias and variance.
Though not applicable in every field, occasionally statisticians run into problems with small sample sizes. Experiments with small sample size tend to be more unstable and biases, resulting in poor parameter estimations. These same concerns apply to experimental designs. This paper investigates the "properties of designs for small experiments" and focuses on minimizing the mean squared error in small experiments.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment