Research Techniques: Measuring Driver Importance
Thursday, October 23
2-3 PM ET
Many market research studies measure desired outcomes, such as brand favorability, loyalty, or wallet share.
This presentation will address various methods to investigate what drives those outcomes and to quantify the impact of each driver.
This session will be non-technical, and it does not require any knowledge of statistics. It is intended to help researchers understand the basics of why driver analysis is done, and why certain approaches are used.
Topics addressed will include:
- Asking importance directly vs. statistically deriving it
- What is correlation?
- The problem of collinearity
- Multiple regression
- Advanced techniques
This webinar is ideal for: Entry and mid-level researchers, project managers, account executives and new analysts who want to better understand some of the most commonly used research techniques.
Gary Parilis, Chief Research Officer, Greenwich Associates
Gary has 18 years professional experience in advanced methods for market research, both on the provider side and as a client. He has led countless analytics projects for businesses of all sorts and in many industries, though the first half of his professional career was focused in technology, and for the past several years he has specialized in financial services. Currently, Gary is Chief Research Officer at Greenwich Associates, leading advanced analytics with responsibility for R&D initiatives. Previously, he spent 10 years at TNS, where he was most recently Senior Vice President in charge of the Marketing Science division, and served as the North American champion for TNS Digital Life, a major global segmentation study on Internet behavior and attitudes. Prior to TNS, Gary spent five years at Lucent Technologies, where he was responsible for analysis and reporting for various market research and customer engagement programs. Previously, he served in teaching and administrative roles at Rutgers University, where he also earned a M.S. and Ph.D. in Social Psychology.
Jeffrey Clarke, Western Governors University
Jeff has over 20 years of experience in statistics and experimental design. He has extensive familiarity with market research analytics and modeling (e.g., segmentation, discrete choice and conjoint analyses, pricing models, brand driver analyses, Structural Equation Modeling, PLS, multiple regression, ANOVA/MANOVA, TURF, decision trees, and data mining). He has managed the analytics for hundreds of projects involving modeling or other advanced statistics. Jeff has conducted advanced analytics in multiple market sectors, including retail, consumer packaged goods, e-commerce, technology, financial services, business-to-business, entertainment, and health care. Jeff has worked in market research both on the vendor side (TNS, MarketCast, Frank Magid Associates) and the client side (Mattel, Inc. and Honeywell), and he is currently an independent consultant. At TNS, Jeff lead the segmentation analysis for the syndicated Global Digital Life study, that included over 70,000 respondents from 60 countries. Jeff has taught statistics and research design in Psychology departments at the University of North Texas and at CUNY Brooklyn College. He earned a Ph.D. in Experimental Psychology from UCLA, and an MBA from the University of Arizona.