Donor Solicitation Dataset

The AnyState Veterans of Foreign Wars periodically solicits donations to fund its various initiatives. To improve the effectiveness of their solicitation campaigns, the VP of Marketing has asked your analytics team to build a predictive model and score their marketing database. Their goal is to spend resources more wisely by targeting those most likely to donate.

The first step is to audit the data that will support this model building effort.

The VP of Marketing has provided four (4) datafiles along with a master data dictionary "Data Dictionary - Donor.pdf." These datafiles are:

Donor_census - as SAS file of census data linked to a donor identification number (CONTROL_NUMBER)

Donor_profile - a csv file with donor demographic data

Donor_history - a csv file with donor donation history data

Donor_survey - a csv file with responses to a donor survey

Overall, there are data on 19,000+- individuals who had been contacted in a veteran's donation campaign in 1997. This file has been pre-qualified for us; there is no further qualification required. The target variable is TARGET_B, which breaks the modeling (qualified) sample into two segments: those who gave during the 1997 campaign; and those that did not. We will explore this dataset over the course of this class.

Task #1 : Using SAS and Word:

Generate a data audit report (using the audit report template as a guide) to be shared with the VP of Marketing; include a check of the available modeling sample size after merging all datafiles and application of target qualifications. NOTE: To support future modeling, your merged dataset should contain one row per CONTROL_NUMBER.

Make sure you review the sample data audit report under Week One Supplemental Material, Variable Annuity Analysis Data Audit (Example).docx, so you understand exactly what is expected for your report. Also use the Data Audit Template.docx as your starting document.

Also, remember the "Questions" section contains questions from you to the marketing department (i.e., What questions do you have based on your audit?) As you cut and paste into the Datafile Details section, make note of data integrity issues you are finding!

Task #2 : Using R:

1. Redo the merging of the 4 "donor" datafiles.
2. Add a new column to the merged datafile called NEW_VARIABLE which has the value of "I did it" for all rows.

Extra

Create a Relational Datafile Structure for the 4 Donor datafiles. See the Appendix to Week 1b Lecture slides for an example (I used PowerPoint). Your structure should show: (i) the name of each datafile; (ii) each file's linking variable name(s); (iii) the number of rows in each datafile; and (iv) arrows which show how the datafiles are sequentially linked/merged with each other.

Homework deliverables:

Task #1:

• Neatly formatted, data audit report (see template) - Word doc.
• SAS program with all code used for the data audit (from SAS Studio, download as a SAS program).
• SAS-merged datafile output as a .csv file labeled with your name and "from SAS", downloaded from SAS Studio.
• Task #2:
  • R program with all code used to merge the 4 donor datafiles and output the resulting merged file.
  • R-merged datafile output as a .RData file labeled with your name and "from R", using save().
• • A word document showing your relational datafile structure graphic.