Data Manipulation and Blending Using Tableau Prep and Workbook

Assessment of Data

We used this data in seminar 6, manipulating and blending data, and we handled survey data in seminar 7. This is very similar but not the same, you have to do some data blending using one different file and the questions you select are different.

You will need to do some data blending and manipulation using Tableau Prep, you will also need to produce a Tableau workbook which addresses the questions asked of you by the MD. You are also expected to produce a report.

In the assessment of the data I will be looking for correct data, correct number of rows and no excess unrequired fields. In the packaged workbook I will be looking for good use of visualisation techniques, a good understanding of the limitations of the analysis and novel application of Tableau functionality. Remember you are presenting this to someone who has not been on the journey of understanding the data with you, so you must tell the story well through the data without confusing or over embellishing.

The final report should be an amalgam of your findings which are sythesised to make final recommendations/conclusions. The suggested report layout is below. The executive summary is your window to the executive decision maker so it is very important to make it clear and concise. It should set the stage for the report by stating the setup, it should use clear diagrams/visuals, and/or infographics. It should state the main findings of the report so make sure you have a sound basis for your conclusions and recommendations. Short paragraphs and bullet points are  examples of good practice on the internet.

Suggested report structure

Digital Business Intelligence reports can take various forms and structures. The structure below is a recommendation but it is not strict. Make your report in MS Word or other compatible programme, a pdf is acceptable. Graphics should be presented from Tableau, if you use others ensure they are licensed under creative commons,

DO NOT USE COMPANY LOGO’s. The final report structure and report organisation is ultimately your decision, however the executive summary and recommendations are compulsory.

Suggested tips for the Tableau workbook;

• Make sure you label each sheet/dashboard with a meaningful name
• If you use an internet tutorial or a book you need to reference it, add a note on the worksheets or dashboard, or include in the report.
• Ensure filters are used correctly
• Explain what is going on clearly and succinctly
• Save your file as a .twbx or I will not be able to mark it properly
• Check everything is doing what it should – sensitivity test
• Use dashboard/storyboards and try to tell a story.
• I will check all Tableau functionality and the data for correctness and ingenuity.
• Apply what you have learned about good visualisation techniques.
• KISS (Keep It Simple Stupid!!)

The Questions & the Data

The data has been gathered at different points in the year and the data engineer originally just sent you the question text with respondent ID’s. You need to amalgamate this data with the appropriate master data to answer the question you have been allocated. These are the files, which you have been given (click on the links and download);

You should explore these files and study their contents, then put the data together in Tableau Prep, as you have been shown in seminars, so that you can get the appropriate analysis in Tableau Desktop. You will need to save a .tde or .hyper data file to use in Tableau Desktop, and the final workbook must be saved as a packaged workbook This format saves your data file along with the workbook, without it the workbook cannot be marked.

The following questions are quite general, this is how you would often be asked to produce analysis in a company. If you have questions about the requirements then I am happy to answer them as if I were your in house client, I will not however, give you a detailed example of what your workbook should look like since you would never get this in reality. You should follow the lead of what we have done in the seminars and use online resources and tutorials. Most importantly you should constantly sanity check your results and try to tell a compelling story with the data even if you get balanced and unexciting results (sometimes that is a good result!). The questions are as follows;

Q1 What effect, if any does geographic location have on the outcomes in the survey? The boss is particularly concerned about work/life balance and wants to present this in the best way possible. Aggregate all of the agree terms and all of the disagree terms and see how this affects your analysis.

Q2 What effect, if any does gender have on the outcomes in the survey? The boss is particularly concerned about work/life balance and wants to present this in  the best way possible. Aggregate all of the agree terms and all of the disagree terms and see how this affects your analysis.