In evaluating new product concepts,practitioners have used aggregate meaures of multi-items and ignore the information value of individual items.Drawing from the Item Response Theory (IRT),this study adopts Differential Item Functioning (DIF) to access bias and impact at the item level.In an online concept testing experiment,study one shows that most items in a multi-item scale display large and significant DIF between major innovations and minor improvements,indicating that different normative benchmarks should be developed for testing major innovations and minor ones.To assess movie popularity using a sample of 329 movies from 2001 to 2009,study two indicates while two different sources of ratings data (consumers versus critics) exhibit moderate difference in discriminating movies popularity,such difference is more pronounced for familiar genres than for those less familiar ones.These findings lend support to DIF as a promising approach to evaluae and select new product concepts and to facilitate effective concept testing and marketing planning.