Premium
Coding of the Independent Variable in Yield Function Calculations 1
Author(s) -
Escobar Giron Jorge A.,
Cady F. B.
Publication year - 1969
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj1969.00021962006100020010x
Subject(s) - coding (social sciences) , mathematics , statistics , confusion , polynomial regression , quadratic equation , shannon–fano coding , variables , variable length code , regression analysis , psychology , geometry , psychoanalysis
Coding or transforming the independent variable(s) in regression situations affects estimation and hypothesis testing. Methods are developed for understanding these coding effects and a table presented with the factors required to calculate easily the estimated coefficients and subsequent t‐statistics in one system of coding from another. The systems of coding include four commonly used codings of the original values. A worked example of a quadratic polynomial model in two variables is then given and the results interpreted with the objective of decreasing the potential confusion that may arise from a coding situation, e.g., when the independent variables have been coded for convenience in calculation but the final yield function is desired in units of the original variables.