How to see the meaning in numbers
2 likes538 views
The document outlines a lesson by Ray Poynter on interpreting and communicating data effectively, focusing on reducing complexity, organizing data, and extracting meaningful insights. It provides methodologies for understanding numerical data and emphasizes the importance of framing business questions to derive tailored narratives from the data. The content includes practical examples, tips on data visualization, and discusses the adoption of automation in data analysis.
How to see the meaning in numbers
- 2. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Agenda 1. Brief recap 2. Reducing Complexity 3. Specific Tips 4. Example with GRIT Data 5. Jigsaw Analogy 6. Top Tips
- 3. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Series Schedule – 2017 • How to see the meaning in numbers 31 May • How to communicate numerical informaKon 5 July • Sense-making with open-ended text 7 September • Think, Feel, Do PresentaKons that change behaviour 9 November
- 4. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter The Frameworks Approach 1. Define and frame the problem – A problem fully defined is a problem half solved 2. Establish what is already known – Find out what is believed and what the expectaKons are 3. Organise the data to be analysed – SystemaKc checking and structural procedures 4. Apply systemaKc analysis processes – Using the business quesKon as a lens to view the data 5. Extract and create the story
- 7. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Making Numbers More Visual • 2 or 3 SD – significant digits • SorKng • Indices (Indexes) • Colour Coding
- 14. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Brand A`ribute Data % Agreeing Sexy Cheap Stylish Friendly Strong Traditional Brand 1 41.73% 63.40% 20.88% 71.20% 41.67% 31.66% Brand 2 87.21% 22.16% 64.50% 40.02% 96.24% 61.31% Brand 3 71.52% 48.80% 53.86% 30.08% 27.88% 28.12% Brand 4 97.23% 9.70% 93.97% 19.89% 96.99% 98.19% Brand 5 92.70% 12.73% 88.36% 34.63% 30.03% 26.98% Brand 6 30.58% 60.82% 20.61% 55.34% 75.00% 64.70% Brand 7 11.47% 72.02% 1.93% 83.76% 59.86% 28.94%
- 15. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Condi-onal Formaang % Agreeing Sexy Cheap Stylish Friendly Strong Traditional Brand 1 41.73% 63.40% 20.88% 71.20% 41.67% 31.66% Brand 2 87.21% 22.16% 64.50% 40.02% 96.24% 61.31% Brand 3 71.52% 48.80% 53.86% 30.08% 27.88% 28.12% Brand 4 97.23% 9.70% 93.97% 19.89% 96.99% 98.19% Brand 5 92.70% 12.73% 88.36% 34.63% 30.03% 26.98% Brand 6 30.58% 60.82% 20.61% 55.34% 75.00% 64.70% Brand 7 11.47% 72.02% 1.93% 83.76% 59.86% 28.94%
- 19. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter What is the Ques-on? • To use analysis to find answers, you need quesKons • To use visualizaKon to find answers, you need quesKons • For example: Does expenditure on health vary by wealth or poli?cs?
- 24. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter 5 Cs of Insight • ConnecKons • Coincidences • CuriosiKes • ContradicKons • CreaKve jump } Build on the current picture Rethink an assumpKon Discard an assumpKon HT, John Storey
- 25. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter GRIT 2017 – Automa-on Example GRIT 2017 Use of Automation Platforms Project design Survey design Sampling Online focus group/IDI moderati on Analysis of survey data Analysis of text data Analysis of social media Analysis of image and video data Report writing Charting and infograph ics Matching suppliers and buyers Matching contract “talent” to projects Analysis of biometric/ nonconsc ious data Attributio n Analytics Not applicable to our organization 27.3% 21.9% 18.4% 24.8% 13.0% 12.5% 21.2% 22.9% 26.0% 12.1% 38.6% 38.5% 34.7% 25.8% Thinking about it but have not tried it 13.7% 14.7% 12.3% 15.9% 13.7% 17.7% 16.2% 17.3% 16.8% 15.1% 10.8% 9.8% 15.6% 11.8% Not sure 29.6% 22.7% 27.3% 23.0% 17.5% 16.9% 16.7% 23.6% 25.0% 17.0% 33.4% 35.7% 24.9% 36.3% Exploring pilot use cases 14.3% 17.8% 17.7% 16.0% 26.3% 31.7% 23.9% 22.4% 17.7% 29.0% 9.3% 9.6% 16.7% 15.1% Already have adopted widely 15.1% 22.9% 24.3% 20.4% 29.5% 21.2% 22.0% 13.8% 14.6% 26.8% 7.9% 6.5% 8.1% 11.0% GRIT 2017 Base, 2637. Report and data will be available 6 June
- 27. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Region and Buyer/Supplier Platform Adoption & Explore / Pilot % North America Europe Rest of World Region Spread Client Supplier Vendor - Client Analysis of survey data 54 58 58 3 53 57 3 Charting and infographics 54 59 58 5 51 57 6 Sampling 42 40 49 9 34 45 10 Survey design 37 44 47 9 35 43 7 Analysis of social media 44 47 50 6 59 42 -17 Analysis of text data 53 54 49 5 53 53 1 Online focus group/IDI moderation 37 35 38 3 38 36 -2 Project design 27 33 31 7 28 30 3 Report writing 29 36 38 9 28 33 5 Analysis of image and video data 34 39 36 5 37 36 -1 Attribution Analytics 26 25 30 6 30 25 -6 Analysis of biometric/nonconscious data 23 29 21 8 26 24 -1 Matching suppliers and buyers 16 17 20 4 14 18 4 Matching contract “talent” to projects 16 15 17 2 16 16 0
- 28. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Region and Buyer/Supplier Platform Adoption & Explore / Pilot % North America Europe Rest of World Region Spread Client Supplier Vendor - Client Analysis of survey data 54 58 58 3 53 57 3 Charting and infographics 54 59 58 5 51 57 6 Sampling 42 40 49 9 34 45 10 Survey design 37 44 47 9 35 43 7 Analysis of social media 44 47 50 6 59 42 -17 Analysis of text data 53 54 49 5 53 53 1 Online focus group/IDI moderation 37 35 38 3 38 36 -2 Project design 27 33 31 7 28 30 3 Report writing 29 36 38 9 28 33 5 Analysis of image and video data 34 39 36 5 37 36 -1 Attribution Analytics 26 25 30 6 30 25 -6 Analysis of biometric/nonconscious data 23 29 21 8 26 24 -1 Matching suppliers and buyers 16 17 20 4 14 18 4 Matching contract “talent” to projects 16 15 17 2 16 16 0
- 29. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Region and Buyer/Supplier Platform Adoption & Explore / Pilot % North America Europe Rest of World Region Spread Client Supplier Vendor - Client Analysis of survey data 54 58 58 3 53 57 3 Charting and infographics 54 59 58 5 51 57 6 Sampling 42 40 49 9 34 45 10 Survey design 37 44 47 9 35 43 7 Analysis of social media 44 47 50 6 59 42 -17 Analysis of text data 53 54 49 5 53 53 1 Online focus group/IDI moderation 37 35 38 3 38 36 -2 Project design 27 33 31 7 28 30 3 Report writing 29 36 38 9 28 33 5 Analysis of image and video data 34 39 36 5 37 36 -1 Attribution Analytics 26 25 30 6 30 25 -6 Analysis of biometric/nonconscious data 23 29 21 8 26 24 -1 Matching suppliers and buyers 16 17 20 4 14 18 4 Matching contract “talent” to projects 16 15 17 2 16 16 0
- 32. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Who is Automa-ng the Most? Question Break High % Medium % Low % Base High - Low % of Total Sample Total All 24 51 25 2637 -2 100 Your professional/ organisal situation Survey Platform Provider/ Survey Software 47 43 10 81 37 3 Number employees 501 - 1000 42 44 13 126 29 5 Research Spending 2017 v 2016 Decrease 38 49 14 168 24 6 Which vertical in? Consumer Staples 42 38 20 132 21 5 Number employees 1001 or more 35 48 17 293 18 11 Will sample get better or worse? Better 34 47 19 680 15 26
- 33. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Who is Automa-ng the Most? Question Break High % Medium % Low % Base High - Low % of Total Sample Total All 24 51 25 2637 -2 100 Your professional/ organisal situation Survey Platform Provider/ Survey Software 47 43 10 81 37 3 Number employees 501 - 1000 42 44 13 126 29 5 Research Spending 2017 v 2016 Decrease 38 49 14 168 24 6 Which vertical in? Consumer Staples 42 38 20 132 21 5 Number employees 1001 or more 35 48 17 293 18 11 Will sample get better or worse? Better 34 47 19 680 15 26
- 34. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Who is Automa-ng the Most? Question Break High % Medium % Low % Base High - Low % of Total Sample Total All 24 51 25 2637 -2 100 Your professional/ organisal situation Survey Platform Provider/ Survey Software 47 43 10 81 37 3 Number employees 501 - 1000 42 44 13 126 29 5 Research Spending 2017 v 2016 Decrease 38 49 14 168 24 6 Which vertical in? Consumer Staples 42 38 20 132 21 5 Number employees 1001 or more 35 48 17 293 18 11 Will sample get better or worse? Better 34 47 19 680 15 26
- 35. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Who is Automa-ng the Most? Question Break High % Medium % Low % Base High - Low % of Total Sample Total All 24 51 25 2637 -2 100 Your professional/ organisal situation Survey Platform Provider/ Survey Software 47 43 10 81 37 3 Number employees 501 - 1000 42 44 13 126 29 5 Research Spending 2017 v 2016 Decrease 38 49 14 168 24 6 Which vertical in? Consumer Staples 42 38 20 132 21 5 Number employees 1001 or more 35 48 17 293 18 11 Will sample get better or worse? Better 34 47 19 680 15 26
- 36. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Who is Automa-ng the Most? Question Break High % Medium % Low % Base High - Low % of Total Sample Total All 24 51 25 2637 -2 100 Your professional/ organisal situation Survey Platform Provider/ Survey Software 47 43 10 81 37 3 Number employees 501 - 1000 42 44 13 126 29 5 Research Spending 2017 v 2016 Decrease 38 49 14 168 24 6 Which vertical in? Consumer Staples 42 38 20 132 21 5 Number employees 1001 or more 35 48 17 293 18 11 Will sample get better or worse? Better 34 47 19 680 15 26
- 37. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Who is Automa-ng the Most? Question Break High % Medium % Low % Base High - Low % of Total Sample Total All 24 51 25 2637 -2 100 Your professional/ organisal situation Survey Platform Provider/ Survey Software 47 43 10 81 37 3 Number employees 501 - 1000 42 44 13 126 29 5 Research Spending 2017 v 2016 Decrease 38 49 14 168 24 6 Which vertical in? Consumer Staples 42 38 20 132 21 5 Number employees 1001 or more 35 48 17 293 18 11 Will sample get better or worse? Better 34 47 19 680 15 26
- 48. Finding and Communica-ng the Story 2017 – Lesson 1 of 4 How to see the meaning in numbers - Ray Poynter Top Tips • Use a frameworks approach • A problem defined is a problem half-solved • Data reducKon is the key to seeing the meaning in the data • Create derived variables, sort the data, use colours • Find the big picture first, then the relevant detail and contradicKons