"Talk Data to Me - the art of storytelling" by Diana Pholo

Slide 1

Slide 1 text

Talk Data to Me: The Art of Storytelling Diana Pholo Predictive Insights

Slide 2

Slide 2 text

Introduction ● Data is just data if we do not effectively visualize and communicate it.

Slide 3

Slide 3 text

Story components ● Situation ● Problem ● Insights ● Next Step

Slide 4

Slide 4 text

The Background

Slide 5

Slide 5 text

Background ● harambee.co.za

Slide 6

Slide 6 text

The Problem

Slide 7

Slide 7 text

Problem Deﬁnition ●

Slide 8

Slide 8 text

Finding the story

Slide 9

Slide 9 text

Exploring the data

Slide 10

Slide 10 text

Exploring the data (cont.) ● Nice tool: pandas ○ head() ○ tail() ○ describe() ○ isnull() ○ etc.

Slide 11

Slide 11 text

Exploring the data (cont.) ●

Slide 12

Slide 12 text

Exploring the data (cont.) ●

Slide 13

Slide 13 text

Exploring the data (cont.) ●

Slide 14

Slide 14 text

Exploring the data (cont.) ●

Slide 15

Slide 15 text

Data Wrangling ● Examples: ○ Convert data types ○ Dealing with NAs ○ Engineer features

Slide 16

Slide 16 text

Data Wrangling (cont.) ● E.g.: Type conversion

Slide 17

Slide 17 text

Data Wrangling (cont.) ● E.g.: Feature Engineering

Slide 18

Slide 18 text

Feature Distributions ● Example: boxplot with Pandas

Slide 19

Slide 19 text

Feature Distributions (cont.) ● Example: Histograms

Slide 20

Slide 20 text

Feature Distributions (cont.) ● Histogram with Pandas

Slide 21

Slide 21 text

Feature-Feature Relationships ●

Slide 22

Slide 22 text

Feature-Feature Relationships (cont.) ● Example: boxplot with Pandas

Slide 23

Slide 23 text

Feature-Feature Relationships ● Correlation

Slide 24

Slide 24 text

Feature-Feature Relationships ● Heatmaps

Slide 25

Slide 25 text

Feature-Feature Relationships ● Barplot

Slide 26

Slide 26 text

Feature-Feature Relationships (cont). ● Scatter plots

Slide 27

Slide 27 text

Real-life example

Slide 28

Slide 28 text

Background ● South African labour market more favourable to men ● Harambee takes in more women than men

Slide 29

Slide 29 text

Problem ● Percentage of women ﬁnding employment is still lower

Slide 30

Slide 30 text

Insights ● Example: Women generally have more responsibilities

Slide 31

Slide 31 text

What’s next? ● Provide child care ● Promote inclusive workplace culture ● Teach shared responsibilities

Slide 32

Slide 32 text

In Conclusion

Slide 33

Slide 33 text

Recap ●

Slide 34

Slide 34 text

Best practices ● Create interesting presentation ● Avoid unnecessary details ● Deﬁne & Understand Audience ● Humanize data

Slide 35

Slide 35 text

Don’t ● Determine narrative beforehand ● Spend more time on presentation than questioning results validity ● Analyse data without big picture ● Use storytelling as a shortcut