To order this White Paper, please submit your email address below:
Deriving Importance for Coded Comments and Categorical Data
Summary of White Paper #7
A while ago, a client called with an intriguing question: “Can you perform key driver analysis using categorical data?” Certainly, key drivers are a germane aspect of survey data analysis; however, many survey attributes are measured on numeric scales such as 1 through 10. Scaled data can be used to calculate distances, which is a prerequisite for common key driver techniques such as correlations or regression. Obviously, categorical data cannot be used to calculate distances – e.g., what is the distance between the categories of “American,” “European” or “Asian?”
To provide additional insight into the data, our client told us that he was currently plotting comments related to problems into a two-dimensional matrix.The first dimension was frequency of occurrence, a value which could be measured quite easily once the comments were manually assigned to a category.The other dimension was the severity of the problem, which was assessed through his industry experience.
A while ago, a client called with an intriguing question: “Can you perform key driver analysis using categorical data?” Certainly, key drivers are a germane aspect of survey data analysis; however, many survey attributes are measured on numeric scales such as 1 through 10. Scaled data can be used to calculate distances, which is a prerequisite for common key driver techniques such as correlations or regression. Obviously, categorical data cannot be used to calculate distances – e.g., what is the distance between the categories of “American,” “European” or “Asian?”
To provide additional insight into the data, our client told us that he was currently plotting comments related to problems into a two-dimensional matrix.The first dimension was frequency of occurrence, a value which could be measured quite easily once the comments were manually assigned to a category.The other dimension was the severity of the problem, which was assessed through his industry experience.