Analysis turns data into information. Our basic analysis customarily includes frequencies and cross tabulations by demographic and respondent characteristics. When more exacting detail is required, we provide population estimates, percentages, standard errors, and confidence intervals around survey responses.
Sophisticated tools are sometimes necessary to get the most information from the data collected. Our staff is skilled and experienced in the use of a full range of univariate and multivariate statistical techniques
This tool finds the underlying construct behind answers to a series of questions. It simplifies the interpretation of responses to many questions into a few key “factors” that drive answers to all questions.
Linear, Logistic, and Logarithmic Regression
These procedures are used to identify predictive variables that can be used to model or predict an outcome. They determine which factors contribute to an outcome and the relative contribution of each factor to it.
Multidimensional Scaling (MDS)/Perceptual Mapping
MDS identifies relationships between entities and objects in terms of how close or far apart they are and describes what determines the distance between them. MDS creates a visual map of the relative position of subjects (institutions, organizations, brands, etc.) using the aspects or features (cost, quality, etc.) that are determined to be most important to respondents. A survey designed to use MDS pairs subjects and then poses questions on whether they are similar or dissimilar with respect to specific attributes. The resulting map provides an easy to understand image of the subject’s position, while ratings of features and attributes identify priorities for action to maintain or change it.
Cluster Analysis uses an iterative statistical technique to group or segment respondents into clusters with similar attitudes, behaviors, and demographic characteristics. It improves understandings of behavior and motivations and allows more effective and precise targeting of marketing efforts.
Discrete Choice Analysis, Conjoint Analysis
Information from discrete choice or conjoint analysis is used to describe the product features (cost, quality, etc.) that are most important for product development, to estimate demand, and to optimize prices. A survey designed for this analysis includes a series of questions; respondent are forced to choose between two options each with a different set of features or “attributes.” The analysis allows modeling to identify the optimal mix of attributes that maximize utility or to drive customer interest.