Analysis of Archetypes

Prototype analysis A woman once talked. But she is more than just an interesting story. She chooses a common image that everyone can understand, puts together a story, and teaches people in this way. . . The storytelling tradition has been in use for a long time as a means to convey wisdom and lessons to other people. One of the most effective ways to achieve this is to use a prototype. You can display these images in the usual way, but you can focus your analysis on a single story.

First, look at the result of prototype analysis. The purpose of prototyping is to identify the extreme patterns in the data so that all observations can be reproduced as a mixture of these extremes. It is easiest to consider the previous sentence, taking physical measurements into account. Look for the shortest person with the basketball team. Look for the person on the top of the basketball team. Everyone in the team is between these extremes. Higher players are more like the highest athletes. Shorter players seem to be the shortest athletes. The height of a particular athlete can be expressed as a mixture of the highest and shortest athletes. Therefore, it can be inferred that the height of 50 to 50 players is between the highest and the lowest half. We add dimensions to the weights and talk about the 2D height, the extreme value in the weight space. I hope that you can understand this.

As long as the underlying structure is defined extremely, prototype analysis is an appropriate model of customer heterogeneity. Therefore, all comparison classifications benefit from prototype analysis. A contrasting classification is used for decisions such as Apple and PC, Democratic and Republican Party, low cost and high end, major banks and regional banks, SUVs and minivans. We exaggerate the difference between products and their users in order to make decisions in the market. Indeed, the desire to use contrast categories in marketing research is so strong that we use K means to identify the clusters and averaging them to analyze these clusters so that the segment names reflect them It will be changed. More extreme prototype members. It means paragraph. Therefore, some users with moderate price sensitivity are marked as having high price sensitivity. Immediately we will forget that this section only shows reasonable price sensitivity