Theme 32: Use of Information and Analytical Systems
Description
Repeated measurements (RM) are common in forage experiments. The data used in this study were accumulated ammonia losses by volatilization (N) and dry matter production (DM) of Cynodon dactylon cv. Coastcross pasture from an experiment in blocks with five levels of urea: 0, 25, 50, 100 and 200 kg of N ha-1, applied in five periods (cuttings). For N, RM were the averages of cuttings and nine days of observation. The F test for the hypothesis of no effect for Period and Level x Period interaction (DM) and for Days and Level x Days interaction was not affected by univariate and multivariate tests. However, Greenhouse- Geisser epsilon estimate was biased downwards. Polynomial contrasts in univariate ANOVA and Logistic function agreed in explaining accumulated N. For DM, unequal population variances on different periods was detected and the assumption that the pairs of observations on the same subject are equally correlated was rejected.
Citation
de Freitas, A. R.; Primavesi, O.; de A. Corrêa, L.; Primavesi, A. C.; Pott, E. B.; and dos S. Mascioli, A., "Repeated Measurement Analyses of Forages in Cropping Systems" (2022). IGC Proceedings (1993-2023). 4.
https://uknowledge.uky.edu/igc/19/32/4
Included in
Repeated Measurement Analyses of Forages in Cropping Systems
Repeated measurements (RM) are common in forage experiments. The data used in this study were accumulated ammonia losses by volatilization (N) and dry matter production (DM) of Cynodon dactylon cv. Coastcross pasture from an experiment in blocks with five levels of urea: 0, 25, 50, 100 and 200 kg of N ha-1, applied in five periods (cuttings). For N, RM were the averages of cuttings and nine days of observation. The F test for the hypothesis of no effect for Period and Level x Period interaction (DM) and for Days and Level x Days interaction was not affected by univariate and multivariate tests. However, Greenhouse- Geisser epsilon estimate was biased downwards. Polynomial contrasts in univariate ANOVA and Logistic function agreed in explaining accumulated N. For DM, unequal population variances on different periods was detected and the assumption that the pairs of observations on the same subject are equally correlated was rejected.
