**Brand Mapping, CHAID, Cluster, Conjoint, Correlation, Factor, and Regression Analysis, Multidimensional Scaling, Structural Equation Modeling, Additional Methodologies
**

**Brand Mapping**** **is defined as the comparison of 2 or more product/service characteristics with those of marketplace competitors. The map is a graphical layout of relative values, resulting in ease of interpretation and comparative ranking with the mean usually at the center and brand product/service distance away from the center representing increasing extremes of the plotted characteristic.

Utilization: Positive characteristic extremes can justify premium pricing while differing characteristic combinations serve to define specific market niches.

** **

**CHAID** (Chi-Squared Automatic Interaction Detection) defines variable interaction by partitioning the target population into distinct groups defined by a set of independent (predictor) variables.

Advantages: Tree display output, easily interpretable visual display.

Disadvantages: Requires larger sample sizes, as small sample size group segmentation may result in datasets insufficient to yield meaningful information.

**
Cluster Analysis** reveals subgroups of the target population exhibiting similar characteristics, as well as differentiation and establishment of subgroups with their own set of similar associations, patterns, relationships, or structures.

Interpretation: The greater the similarity within subgroups and the greater the difference between groups, the more distinct the clustering.

**
Conjoint Analysis** (Trade-Off Analysis) measures respondent perceived values of specific product features, acceptable pricing, and delineation of features the participants are willing to sacrifice in maintaining bundling arrays at various price points.

Example: Computer systems features offering various processor speeds, amounts of RAM, video card quality, hard drive size and speed, as they affect purchase pricing.

**
Correlation Analysis** measures the linear relationship between two variables as described by their correlation coefficient (range of +1.0 - - 1.0 where +1.0 represents perfect positive correlation and direction of change given similar conditions, 0 represents no relationship between the two variables, and -1.0 represents maximal negative correlation or opposite directional change given similar conditions).

.

**
Logistic Regression Analysis** is predictive when the dependent variable is dichotomous (having two possible outcomes, usually one positive and the other negative, for example describing survival or non-survival among those with a particular diagnosis) and the independent variables are continuous and discrete (e.g., defining the diagnosis present in the targeted groups).

**
Factor Analysis**, commonly used in customer satisfaction studies and the Financial Services industry, analyzes large numbers of dependent variables to detect certain aspects of the independent variables (referred to as factors) affecting those dependent variables, without directly analyzing the independent variables. Requires underlying data distribution as multivariate normal, with linear relationships. A correlation matrix must first be calculated prior to performing factor analysis.

Utilization: The description of many variables and their characteristics utilizing fewer factors.

**
Multidimensional Scaling** (MDS) is used to visualize dissimilarity data, such that the objects are represented as points usually in a two dimensional space, where the distances between the points match the observed dissimilarities as closely as possible. Data is less subject to the restrictions of factor analysis, and can be used directly and applied to any kind of similarities or dissimilarities.

**
(Linear) Regression Analysis**, or **Regression Modeling**, is used to quantify a best fit (usually represented by a straight line on a scatter plot) between the dependent variable (i.e., the variable researchers are attempting to predict), and the independent variable(s) that are chosen for evaluation (i.e., the known or hypothesized predictor variable(s)). Multivariate linear regression analysis is utilized to model two or more continuous or categorical predictor variables and a continuous outcome variable.

Output: Correlation coefficient (see above), determination coefficient (measure the proportion of variation in the outcome variable explained by the regression model, and ANOVA (analysis of variance), a gauge of linear association between the outcome and predictor variables.

**
Structural Equation Modeling** includes several multivariate analysis methods and is applied in the setting of causal modeling, confirmatory factor analysis, second order factor analysis, regression models, covariance structure models, and correlation structure models.

Output: Easily interpretable path diagrams.