RAM NIWAS*, VINAY KUMAR, O. P. SHEORAN AND SHEETAL CHAUDHARY
Department of Mathematics and Statistics
CCS Haryana Agricultural University, Hisar-125 004 (Haryana), India
*(e-mail: rniwas@hau.ac.in)
(Received : 3 March 2024; Accepted : 25 March 2024)
SUMMARY
This research paper investigates the complex relationships influencing pearl millet yield using Structural Equation Modeling (SEM). The study employed data from pearl millet genotypes, encompassing various yield components and attributing traits. Principal Component Analysis (PCA) was utilized to identify latent variables representing dimensions related to yield determination. A recursive SEM with latent variables was constructed, incorporating measurement models for exogenous and endogenous variables and a structural model depicting causal relationships. Maximum likelihood estimation was used to estimate model parameters, and goodness-of-fit criteria were applied for model evaluation. The study revealed fertility’s pivotal role in determining pearl millet yield, with tillering capacity and panicle weight identified as crucial factors. Physiological and morphological traits also influenced yield but to a lesser extent, indicating a complex interplay of factors. The final SEM model provided valuable insights into these dynamics, highlighting the intricate relationships and pathways involved in yield determination.
Key words: Pearl millet, yield determination, Structural Equation Modeling (SEM), Principal Component Analysis (PCA), latent Variables, exogenous variables, endogenous variables, fertility, tillering capacity, panicle weight