Overview
Model the sales of one brand at one retailer, using a regression model, based on the secondary sales dataset used in the ‘introduction to R’ session. You can choose any brand/retailer combination you like. For example, if you choose brand 1 at retailer 1, your task is to build an understanding of sales drivers for brand 1 at retailer 1. You do not need to motivate this decision, but it would be good to ensure there is some variation in the data to model.
It is up to you to identify sales drivers. You can manipulate the data in any software you like (R, Excel). In estimating the model, please use R and submit your code along with the report. If you manipulated data outside R, please also submit the Excel file. If we run the submitted code on the submitted Excel file, we should be able to replicate the model results presented in your report.
You should clearly present 1 preferred model that forms the focus of your report, but you can report additional models to motivate your preferred model and/or to illustrate the robustness or validity of your preferred model.
Format
Times New Roman 12, single spaced, maximum 6 pages (excl. appendices, cover page, table of contents, reference list).
Evaluation
Your submission will be graded according to the following criteria:
Modelling application (motivation and clarity of the proposed modelling, use of available data and modelling techniques to build an understanding of sales drivers, interpretation of results and consolidation into conclusions and insights) (50%)
Critical analysis (quality of critical thinking on the modelling and its validity, quality of limitations and potential solutions) (30%)
Communication (quality of presentation of the report, appropriate use of figures/tables and their integration into the main part of the report, clear logical flow and consistency) (20%).