The App Happy Company wants to better understand what they believe is the market for a new social entertainment app they are thinking of developing. They are currently in the business of providing B2B analytic apps, and don't have yet have a product in the consumer entertainment app category.
App Happy hired the Consumer Spy Corporation (CSC) to survey consumers in the market. CSC collected data from a sample of consumers, and provided App Happy with a dataset of of their responses. The survey questionnaire (it's in the file apphappy-quex-sum2014.pdf) was based on some preliminary qualitative research that included focus groups and one-on-one interviews.
(1) App Happy wants you to use their survey data to do a general attitudinal post hoc segmentation analysis. You should develop and evaluate a segmentation scheme for App Happy, profile the segments in your scheme, interpret your results, and make recommendations about product opportunities or additional research as might be appropriate. Be sure to describe your analytic methodology and to summarize assumptions, caveats and limitations that App Happy should bear in mind in using what you report to them.
(2) App Happy would also like to use your results to put consumers that it doesn't currently have data on into the segments you will define for them should it be able to obtain data on them. It would like you to describe how you would develop a classification model that could do this. You don't have to actually apply the model to get results. You just need to describe what model or models they could use, and provide enough explanation that App Happy would be well informed to apply the model(s) you recommend by itself. In the application domain of segmentation analysis, such models are sometimes called “typing tools.” (But they have nothing to do with keyboards...)
You may use additional information or data about the market as seems useful or important to you.
Be sure to check the data for errors or anomalies as a preliminary step in your segmentation analysis. Also, consider whether data transformations may be useful. Note that the data are real data and the documentation on them is less than perfect. Finally, be sure to address all the above issues and questions.
App Happy's data are in the R data file apphappyData.Rdata. This file contains to R data frames: apphappy.3.num.frame, and apphappy.3.labs.frame. The first of these has numerically coded survey response data. The second has response data coded in the character strings of the questionnaire's value labels. These strings are like the value labels in SPSS, or the labels you might find attached to numerical codes in SAS when Proc Format is used.