Think Like a Data Scientist: My 168 Hours in a Week

This assignment was inspired by the article How to Start Thinking Like a Data Scientist by Thomas C. Redman in the Harvard Business Review. (The link takes you to the article online, but its not essential that you read itmore of an FYI if youre curious.)

You dont have to be a data scientist or a Bayesian statistician to tease useful insights from data. For 20 years, Ive used a simple exercise to help those with an open mind (and a pencil, paper, and calculator) get started. One activity wont make you data savvy, but it will help you become data literate, open your eyes to the millions of small data opportunities, and enable you to work a bit more effectively with data scientists, analytics, and all things quantitative. Each step in this assignment illustrates an important concept in analyticsfrom planning for data collection to understanding variation to data visualization and clear communication.

Objective(s)

Demonstrate how to apply the data analytics process from initial question to reporting

Communicate each step in the process clearly and concisely using data storytelling techniques in the Slidedoc format

Instructions

Each of you will answer the same basic question: How do I invest my time in a week? Because everybody has an equal amount of time168 hours (24 hours/day x 7 days)but uses it in different ways. Hopefully, it will be insightful to analyze what a typical week looks like for you.

Start by thinking about your life and how you use your time to identify your main categories and sub-categories. Most of you will have a minimum of four main categories and depending on your lifestyle possibly one to two others. Theres no right/wrong answer hereit depends on whats important to you.

Main Category	School	Me Time	Family	Sleep	*Other
Sub-Categories	Classes Homework/ Assignments Studying Admin	Watch Listen Read Hobby Fitness Sport	Meals Chores Time w/partner Time w/kid(s) or Kid stuff	Zzz	Part-time job Pet Hobby Fitness Sport

Situation Overview

Paint a picture (so to speak) of your lifestyle to help the audience understand how youre approaching the question based on your specific situation.

Describe your current situation to give background/context

Identify your main categories

Based on your situation, identify at least two of the categories youre curious to investigate more in-depth (using the sub-categories) based on an assumption and/or hypothesis you have

Data Collection Plan

Outline the specific plan to get the data you need to answer the question. Use the correct terminology.

In a table (like above), identify the categories and sub-categories of data you are going to capture as individual data points

Define any data points that might need definition

Recommend working in 15 to 30 and 60 minute increments

Describe how youre going to collect the data and the timeframe

Identify data sources used to collect data (i.e. pen/paper, smartphone, tracking app etc.)

If necessary, briefly explain if you had to adjust the original plan after you started to collect the data

Raw Data Set

Organize and display the raw data you collected in a table

Dont get so complicated that the data cant clearly be shown

Show data for each day and time increment

Use a legend to colour code each category if it helps

This should be raw data so no ranking, totals, percentages, graphs at this point

If its more than 1-page take a screen snip of a representative part of the table to show how youve organized it, and include the rest of the raw data in an Appendix

Describe your confidence in the quality of the data you collected

Is the week you picked representative of a typical week?

Did you diligently capture the data as planned?

Analyze the Data

Doing analysis means combining, organizing, ranking the data in a way that makes sense for the questions to uncover relationships and results.

Analyzes the main How do I invest my time in a week? question

Analyzes (at least) two categories in-depth that youre curious about (based on your assumption or specific hypothesis)

Explain (tell) what you are seeing in the datai.e. patterns, trends, outliers

Create data visualization (show) like bar graph and/or pie chart to complement the narrative

Evaluate what youre seeing in the analysis based on the important questions. What insight(s) emerge?

Identify the main insight(s)/ah-ha!/takeaway from each analysis

This shouldnt be a long list for each analysis, its the most important one to two things (that could help you moving forward)

Use one page for each analysis (main 168 hours, in-depth 1, in-depth 2 etc.) that applies each of the typical Slidedoc elements for a page

Recommendations

This is where the analysis and insights come together. How can you action this/use this in the future? Where do you go from here?

Make four (4) recommendations for yourself:

Continue to keep doing something thats working

Change something youre doing to make it better

Stop doing something thats just not working

Start doing something new

Things to Consider

Situation Overview, Data Collection Plan, Raw Data Set, Recommendations can each be done in 1-page.

Analyze the Data will have 1-page for each question or category youre investigating.

Assignment Format

Create a multi-page report using one of the Slidedoc templates (in Microsoft PowerPoint) that was provided in the course resources

Aside from cover, table of contents and appendix; each page of a Slidedoc is organized to include these elements along with white space:

TITLE

Main takeaway

Supporting body copy

Data visualization (usually 1, absolute maximum of 2 on a page)

Make sure each element on the page is clearly readable at 100% page size ("zoom")

Typically, this means the minimum font size of 12pt for the important elements

Submit the final document in PDF formatwhich will keep your original formatting intact