Storyboarding Your Data Story
Big picture -> details -> big picture
- Why? So what?
- SMART rec. (1 minute)
- What I am going to show
- Supporting evidence
- Link evidence to SMART rec.
- What you could gain
- Why? So what?
Making your data story come alive
Bring the pride to your presentation.
Closing in an emotional way.
- Actual story, picture in their mind
Tell a real life story to show the story elements
- Family grouping dinner
- Show the direction and satisfaction from the story
- Let the audience the feel the same emotion
To start the presentation
- Use an image
Find something that is motivated to you!
- Specific details
- Positive messages -> positive emotions
- Use large, high-quality pictures
Stress-testing Your Story
Present the data story to your stakeholders.
Team culture: the culture of bring forward any errors and that being celebrated.
Identify inconsistencies and logical arguments
Choose a class in Logic and Reasoning in Couresa
Overgeneralization and Sample bias
Assume what you are seeing in your dataset is what you would see.
- Your data is very small or sometimes it’s selected subset (biased).
- A lot of missing data in your data set
It is tempting to think that if you just have a big enough data set, you should be able to overcome most types of sampling bias.
Smartphone owner: wealthier and younger people
Poor decisions and inaccurate predications.
Data shadows: lack of data
- It can take a lot of detective work to figure out that your sample is biased in the first place
Single-mindedly paranoid about your data quality is actually one of the biggest ways that you can prevent mistakes.
How to avoid
- Ask a lot of questions about how your data was collected
- Check how many data points you have in all the groups you’re looking at
- Split your full data set into three to five random subsets. See if you observe the same effects n each one as you see in the group as a whole
Misinterpretations Due to Lack of Controls
Include designed comparison groups that should not have the effect you were looking for in your analyses to make sure the effects you observe are due to the events you think they are due to.
Correlation Does Not Equal Causation
Refers to the phenomenon of two things having a tendency to vary together over multiple time points or multiple measurements.
Refers to the phenomenon of one thing happening as the result of the other thing.
When purely by chance two things happen at the same time.
When either due to chance or due to an unmeasured variable, but not by direct causality, two tings correlate.
The percentage of users who click on an advertisement link when they see it.
How Correlations Impact Business Decision
Change the variable you think is causing the effect you want and hold all other things constant.
Test on two different groups
- Every time you see a correlation between two entities related to a business recommendation you wanna make, get in the habit of questioning whether there is some other third or fourth or fifth variable that can explain the relationship you see.
- Then look for data that allows you to test whether that third or fourth or fifth variable is a better measure of the phenomena you are interested in.
- Examine whether the correlation you’re basing your business recommendation on exists in other contexts or datasets. The more you can replicate the effect, the less likely the first correlation you saw was due to random chance.
- Try to come up with different, but complementary angles to assess the causal relationship you’re hypothesizing about.
The likelihood that you will get into trouble inferring causation from correlation, increases as the size of your data sets increase.
Likelihood for getting into trouble increase as the complexity of your data sets increases as well.
When you don’t know why two phenomena are correlated, you don’t know how to predict when their correlation might change.
Data is meant to inform human decision making, not replace it.
Think of data as a resource to increase the number of good decisions made.
Tools for Conveying Your Data Story
Use bar charts and line charts as defaults
Purpose is to convey the most critical parts of your data.
Decision making process
Used for comparing measures in different groups or categories.
Show aggregated rather than raw data
How values and categories vary over time.
Specific, inherent, sequential order
Show raw data
Best fit line
Too confusing and overwhelming for non-technical audiences
- Bar charts (comparing categories)
- Line charts (changes over time or an ordered category)
- Pie charts (4 or fewer categories that add up to 100)
- Scatter charts (unless you have a technical audience)
- 3D charts (ever)
Relative differences in position and length
Our eyes and brains don’t have the ability to perceive many of the differences other charts rely on.
- Positions on a common scale
- Angles and slopes
- Color saturation or shading
- Color hue
Misinterpretation Caused by Colorbars
Do not use color to
- Convey detailed quantitative differences in the values of continuous variables
Do use color to
- Illustrate general patterns
- Code for different categories of categorical variables
- Draw attention to something
Visual Contrast Directs Where Your Audience Looks
Influence what they look at.
You have already decided what parts are important.
And you goal is to only show your audience they need to make a decision about your recommendation.
Visualization for data analysis should show as much data as possible.
Visualization for persuasion should show selected pieces of data and should direct your audience’s eyes to the precise points of the data that support the arguments you are trying to persuade them of.
Stand out relative to their neighbors.
The human brain can only handle a limited amount of information at a time.
Our brains developed a focus on important things in the environment and filter out everything else.
Make sure they see only the parts they need to see in order to make a decision about your recommendation.
Putting Compelling Data Visualizations into Persuasive Business Presentations
Formatting Slides to Communicate Date Stories
Maximize the Data-Ink ratio
Data ink: ink that represents the actual data
Non-data ink: everything else
No extra information and text on the page
Understanding at a glance
Put label on the bar charts
Times new roman, bodoni, garamond
Helvetica, calibri, arial
Don’t make your audience do visual math
Split the complex charts
Formatting Presentations to Communicate Date Stories
In 10 minutes
- 3 dividend, left 2 is title, right 1 is picture
- Check for typos. Then check again
- Bold is readable than italics and underlining
- No distorted or fuzzy pictures
- Use 2-3 colors (but others can be used for highlighting)
Delivering your data story
What you will say
How you will say them
Fast, slow speed