Application of logistic regression statistical technique to evaluate tomato quality subjected to different pre- and post-harvest treatments

Abstract

The study applied a combination of categorical and continuous explanatory variables via logistic regressions to model the probability of quality, defined as marketability, of tomato (Solanum lycopersicum L.) fruits subjected to different pre-harvest (biocatalyst application) and post-harvest treatments: (i) dipping in tap water, chlorinated water or anolyte water (ii) thin film packaging versus no packaging, and (iii) storage at 13 °C or in ambient temperature. Fungi, bacteria, coliforms, total soluble solids (TSS), glucose, fructose and ascorbic acid were determined in a 30-day full factorial experiment. The effects of each quality variable on the marketability of tomato fruits were explored. A multiple logistic regression model, consisting of both the continuous and the categorical variables, was then derived to evaluate the effectiveness of the anolyte water as dipping treatment. The simple logistic models showed that storage time, fungi, aerobic bacteria and coliform population are negatively related to the probability of marketability. In contrast, TSS, glucose and fructose are positively related to the probability of marketability. No significant relationship was observed between the probability of marketability and ascorbic acid content or pre-harvest biocatalyst application. The multiple logistic regression model showed that the probability of marketability was higher when the tomatoes were subjected to anolyte dipping treatment, packaging and storage at 13 °C. The study presented new insights into the interpretation of post-harvest quality defined as probability of marketability of the stored fresh produce.