Nonparametric estimation of bivariate additive models

Quality Assessment of Greek-Style Set Yoghurt Fortified with Extracted and Dried Australian Native Fruit, Illawarra Plum

The Illawarra plum (IP) is native to Australia and has been used as a bush food for centuries. With rich phytochemicals and natural pigments, IP has the potential to be an added ingredient to improve the physicochemical properties of food, including yoghurt. This study prepared dried IP powders through vacuum drying (VD) and freeze drying (FD), produced extracts from these dried powders, and fortified them into Greek-style set yoghurt. The changes in colour, pH, titratable acidity (TA), syneresis, total soluble solids (TS), and phytochemicals were measured throughout a chilled storage period of 28 days. The results showed that FD and FD extract could provide a stable, distinct pink colour to yoghurt. IP powders and their extracts improved TS content and, thus, the consistency of yoghurt. Compared to the control, VD, FD, and FD extract of IP did not significantly affect the level of acidity or syneresis after 28 days of chilled storage. Yoghurt fortified with FD and FD extract had the greatest level of phenolics, anthocyanins, and radical scavenging antioxidant activities. This study revealed that IP powders and their extracts can positively improve the physicochemical properties of Greek-style set yoghurt. FD powder is recommended over its extract to fortify this yoghurt, as it can be cost-effectively prepared by freeze drying and then grinding the fresh fruit into powder. Future studies are needed to evaluate other variables in yoghurt preparation, including the concentration of IP and strains of yoghurt culture. Further, the consumer perception of sensory quality and acceptability of yoghurt fortified with FD IP powder should be investigated.

Backfitting tests in generalized structured models

We introduce bootstrap tests for semiparametric generalized structured models. These can be used for testing different kinds of model specifications like separability, functional forms and homogeneity of effects, or for performing variable selection in a large class of semiparametric models. The test statistics are based on the comparison of non- and semiparametric alternatives in which both the null hypothesis and the alternative are non- or semiparametric. All estimators are obtained by smooth backfitting. Simulation studies show excellent performance of the test procedures.