Applications of Rasch analysis in consumer research for new food product development

Introduction
The classical test theory (CTT) and its extensions are the predominant measurement and data processing approaches in current consumer research areas. However, the CTT-based approaches suffer from several theoretical drawbacks such as requiring interval data and lack of quality control procedures. To overcome these drawbacks, an alternative method Rasch analysis, which stands for the analysis using a family of parametric probabilistic response models, can be used. Although Rasch analysis has been broadly used in education and health research areas, it is yet to be used extensively in food-related research and new food product development.
The aims of this research were to demonstrate the benefit which can be obtained by applying Rasch analysis to consumer research for new food product development and to explore the application of Rasch analysis in the development and validation of an instrument which can be used in food-related consumer research tasks

Research activities and outcomes
To achieve the aims of the research, four case studies were conducted in the following order:
Case study I compared the difference between CTT approach and Rasch analysis in the evaluation of a survey of 269 respondents using an existing instrument Health and Taste Attitude Scales. The results indicated that compared to CTT approach, Rasch analysis could identify more meaningful underlying structure of the instrument and interpretable reliability statistics.
Case study II employed the Many-Facet Rasch model in a sensory study for modelling a composite overall liking measure from 8 sensory attributes and a holistic measure from a single overall acceptability item. The psychometric properties of the two models were compared. The ability of the two measures to differentiate between the overall liking of product were also compared. The results suggested that the composite measure has greater ability of differentiating products.
Case study III developed and validated a set of instruments in relation to ready meal consumption under the guidance of Rasch analysis. The instruments can be used for measuring consumers’ satisfaction attitudes, decision making patterns and willingness to consume ready meals in different contextual situations. The hierarchical rank order of the items also provided information associated with new food product development opportunities.
Case study IV developed and validated a sensory instrument for benchmarking test on beef lasagne ready meals under the guidance of Rasch analysis. The result verified the expectation of product ranking. The information also reflected panellists’ particular needs for product improvement.

Conclusion
Rasch analysis can overcome the limits of CTT approach, improving the quality of measurement in consumer research for new food product development practice. It should be applied to more food-related area.