Premium
Bias Error in Yield Functions as Influenced by Treatment Design and Postulated Model
Author(s) -
Cady F. B.,
Laird R. J.
Publication year - 1969
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj1969.03615995003300020033x
Subject(s) - variance (accounting) , statistics , variable (mathematics) , selection (genetic algorithm) , selection bias , econometrics , mathematics , yield (engineering) , computer science , economics , mathematical analysis , accounting , artificial intelligence , materials science , metallurgy
Bias error occurs when the fitted or postulated model is not equal to the true functional relationship between the dependent variable and the independent variable(s). The concept of bias error is developed and the effect of various factors on magnitude of the bias error studied. The factors included in an empirical study are spacing and number of fertilizer levels and selection of the postulated model. The choice of a postulated model, a factor not under the control of the experimenter during the planning stages of an experiment, shows the largest influence on bias error. The selection of the postulated model is discussed and suggestions for the control of bias and variance error in response surface experimentation are given.