Effects of Initial Conditions on Monte Carlo Estimates of Bias in Estimating Functional Relationships
- 1 January 1988
- journal article
- research article
- Published by Canadian Science Publishing in Canadian Journal of Fisheries and Aquatic Sciences
- Vol. 45 (1) , 185-187
- https://doi.org/10.1139/f88-020
Abstract
Some of the results presented by Walters (1985. Can. J. Fish. Aquat. Sci. 42: 147–149) for the magnitude of bias in estimating functional relationships from time series data resulted from his choice of initial stock size in Monte Carlo simulations rather than the dynamics of the model. Walters used the same initial stock size in each simulation while varying parameters in the stock–recruitment relationship. Starting each simulation at the equilibrium stock size or allowing initial stock size to vary randomly produces larger estimates of bias and leads to different conclusions about the relationship of bias to parameter values in the model.This publication has 2 references indexed in Scilit:
- Factors Affecting the Time Series Bias in Stock–Recruitment Relationships and the Interaction Between Time Series and Measurement Error BiasCanadian Journal of Fisheries and Aquatic Sciences, 1988
- Bias in the Estimation of Functional Relationships from Time Series DataCanadian Journal of Fisheries and Aquatic Sciences, 1985