Transcript PRESENTATION NAME
A Geography of Taste: Coffee Quality and Spatialized Tag Clouds
Frank LaFone, Bradley Wilson, Trevor Harris Department of Geology and Geography West Virginia University
The $13 Cup of Coffee
• “It’s very fruity, juicy, good mouth feel, [and] full bodied.“ Jay Caragay, Baltimore coffee shop Spro, selling a 12 ounce cup of for $13 • "The cup was very balanced, with slight orange notes. It had a clean finish… In the start and through the finish, in the aroma and in the taste, you can identify many different notes."“ Caroline Bell, Brooklyn coffee shop Café Grumpy selling a $9 cup of coffee • “How you brew isn't as important as what you brew…So it's all about the bean.” Alan Kaiser of the National Coffee Association • “ $13 coffee worth the brew-haha?
” by John DeVore and Steven Stern, CNN April 7, 2010
Background: The Coffee Paradox
• Latte revolution - Dramatic changes in coffee procurement and retail marketing since 1980s • Buyer-driven trade dynamics - small, medium and large-scale roasters and retailers w/ increased control over value chain • A ‘coffee boom’ in consuming countries and a ‘coffee crisis’ in producing countries
Background: Existing Geographic Research
• Global Value Chains, Economic Restructuring, Crisis • Agro-Ecology of Farming Communities, Biodiversity Conservation, Livelihoods • Coffee Certifications, Market-based Development Initiatives • Ethical Consumerism, Alternative Trade Networks • What about Taste and Place?
Taste and Place
• Taste is subjective, personal, complex – yet taste is used to determine coffee quality and thus price • “special geographic microclimates produce beans with unique flavor profiles” (Knutsen) • Taste continues to trump other qualities (i.e. ethical, ecological values) – Sensory Evaluation = Buying Decision – Professional Coffee Quality Evaluators • Try to identify unique flavor profiles of coffee produced around the world • Enforce of quality standards
Defining Specialty Coffee
“Until the moment that the roasted coffee is brewed and transformed into a beverage, the concept of specialty coffee is locked up as a possibility, just a potentially wonderful gustatory experience. Starting at ground level, so to speak, we must limit specialty coffee to those that are drawn from the appropriate intersection of cultivar, microclimate, soil chemistry and husbandry.” - Ric Rhinehart www.scaa.org
Coffee Cupping
Cup of Excellence
• Quality Competition and Internet Auction • “The Oscars of Coffee” – “Highest Echelon” • Started Brazil 1999, Guatemala 2001 • 2010 – Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica, Colombia, Brazil, Bolivia, Rwanda – Most stringent coffee cupping standards and evaluation techniques – 300-500 competition entries /country /year
What makes an “excellent” cup of coffee?
• Sought to understand Cup of Excellence competition and auction outcomes • Spatial analysis of Central America results 2003-2009.
– Back link coffee evaluation from cup to crop • Explore environmental factors effecting taste - varietals, soils, climate, and farm practices, regional geographic location (terroir)
Database Development
• Barrier: No data stored in traditional database forms or formats • Recorded information on paper, translated into website ( http://www.cupofexcellence.org/ ) • Only “winners” scoring above 80 recorded • Information copied from the website and placed into an initial database for analysis
Cup of Excellence Winners By The Numbers
Countries Years Number of Winners Minimum Score Maximum Score Median Score Mean Score Minimum Price /lb.
Maximum Price /lb.
Median Price /lb.
Mean Price /lb.
Nicaragua, El Salvador, Honduras, Costa Rica, Guatemala 2003 – 2009 756 80.25
95.76
85.95
86.70
$1.16 $68.17 $3.83 $4.61
Cup of Excellence Competition Data Description – Environmental Variables
Variable Range, Examples
Year 2003-2009 Country City Region Coffee Variety Processing Method Farm Name Farmer Farm Size Coffee Growing Area Altitude Certifications Honduras, Costa Rica, Nicaragua, Guatemala, El Salvador Citalá, Jinotega, etc Corquin, San Juan Sacatepéquez, etc Bourbón, Caturra and Catuai Washed and sun dried, Conventional, etc Santa Elena II, Shangrilá, etc Julia Rosa Mena de Lima, Ernesto Lima Mena, etc Hectares Hectares Meters Fair Trade, Organic, etc
Cup of Excellence Competition Data Description – Competition Variables
Variable
Rank Score Lot Size Cupping Number Price 2003 Dollars Winning Bidder Descriptors
Range, Examples
1 - n (n differs by country/year) 80-100 Number of bags/boxes submitted Unique ID number USD in current year Derived from above World Coffee Co.,Ltd., The Rosterie Inc, etc.
Citrus, creamy, mellow, complex, chocolate, sweet, floral, melon, butterscotch, rich, orange, carrot, jasmine, honey, spicy, syrupy, licorice
Exploratory Spatial Data Analysis Quantitative Approach
• No notable correlations for environmental factors (e.g. altitude vs. price).
• Notable correlation price (i.e. auction outcome) vs. rank (i.e. competition outcome) – Clear outliers – Variability in each rank – Why do certain entries get better auction when the get lesser rank?
– Does flavor (descriptors) drive price?
Exploratory Spatial Data Analysis Qualitative Approach • Created Dictionary of Coffee Descriptors – Massaging the database • Flavor wheel too limiting • Descriptor groups by country/ rank/ price/ etc • Need for qualitative analysis – flavor descriptors influence the outcomes
Approach To Qualitative ESDA
• Visual tools can help the brain efficiently process complex ideas and relationships • Wordles – Created by Jonathan Feinberg of IBM – Owned and copyrighted by IBM – Display words in different sizes based upon word counts in a piece of text – Outputs are designed primarily to be an ‘artistic’ rendering, not scientific analytical – Can compare wordle outputs to see differences in word counts in two (or more) pieces of text
Wordles
Comparative Wordles
Highest Dectile of Rank Compared to Lowest Dectile of Rank
Comparative Wordles
Highest Dectile of Price Compared to Lowest Dectile of Price
El Salvador Comparative Wordles
Highest Dectile of Rank Compared to Lowest Dectile of Rank
El Salvador Comparative Wordles
Highest Dectile of Price Compared to Lowest Dectile of Price
Nicaragua/El Salvador Comparative Wordles By Rank
Future Directions
Conclusions
1.
2.
Environmental variables aren’t important – anyone can win Flavor wheel isn’t that important 3. Complex dictionary of descriptors Semiotically rich 4. Difference in descriptors for rank and price 5. Differences between countries 6. Descriptors are non-random they have notable outcomes
Questions? Comments? Thank You
Frank LaFone, Bradley Wilson, Trevor Harris Department of Geology and Geography West Virginia University
Nicaragua Comparative Wordles
Highest Dectile of Rank Compared to Lowest Dectile of Rank