Improved NIRS Calibration for Hardness, Resilience, Cohesiveness, Springiness and Adhesiveness of Gari/Eba (cooked dough), at IITA, Ibadan, Nigeria.

The near-infrared reflectance spectra of Gari_Eba produced from 23 cassava genotypes were collected using the XDS-Rapid Content Analyzer. Gari_Eba dough was homogenized and filled into the equipment's ring cell sample holder. NIR spectra were collected in duplicate for each sample in the 400 to 2485 nm wavelength range. Before NIRS analysis, the instrumental texture profile of each sample was determined using the TA.XT texturometer and the same sample lots were used for the NIRS measurements. Reference data for hardness, cohesiveness, resilience, adhesiveness and springiness were obtained and combined with the NIRS spectra to develop prediction models for the textural attributes. Improved prediction models for the textural qualities of Eba were obtained by pooling the dataset from 2022 and 2023. The coefficient of determination and prediction ( R2cal and R2pred) for hardness, resilience, adhesiveness and cohesiveness were ( 0.83 and 0.65), (0.73 and 0.69), (0.71 and 0.23), (0.79 and 0.43) and (0.77 and 0.50) respectively.