D'Amico Associates
Experts in Consumer Insights & Market Research
HomeAbout UsThe ProgramQuestionsFallaciesStatistical ApproachesContact Us

Statistical Approaches

To circumvent many of the traditional media and market research fallacies and to provide users with a better understating of the research data, the Consumer Insights & Charting Program uses various sophisticated analytical and statistical approaches. These include:

  • Target association coefficients to determine how well a given measure helps predict brand usage
  • Discrimination coefficients to determine which measures best differentiate any two user groups or brands
  • Incidence Indices (based on odds ratios) to determine a segment's relative propensity to use a given brand or service
  • Explained variance to evaluate all measures on an equal footing in terms of their ability to predict brand usage
  • Compositional differences to determine a brand's competitive strengths and weaknesses
  • T-tests to determine whether brand differences are statistically significant 
  • Statistical significance evaluated at seven different levels of confidence
  • 95% confidence limits to determine the variability surrounding group or brand differences
  • Brand rankings to determine how well a brand preforms relative to other brands in the analysis on a given measure relative to both audience size and index

To find out more about the flawed assumptions that underpin many traditional media and marketing approaches, click on the links below:

D'Amico, T., "A New Approach for Profiling Brands And Analyzing Competitive Information", Quirk's Marketing Reseach, August 2011.

D'Amico, T., "Magazine's Secret Weapon: Media Selection On The Basis of Behavior, As Opposed to Demography", Journal of Advertising Research, November/December 1999.