Letting Data Speak Enunciative Modalites of Correspondence Analysis - PowerPoint PPT Presentation

Letting Data Speak Enunciative Modalites of Correspondence Analysis Email: Richard@Volpato.net Richard.Volpato.net

Letting Data Speak -- the Truth Veritas Filia Temporis Truth is the Daughter of Time From: Francesco Marcolino 1536, Venetian printer Richard.Volpato.net

Letting Data Speak -- the Truth Veritas Filia Temporis Truth is the Daughter of Time From: “Statistics is not a means of Francesco Marcolino knowledge, draining into an idea 1536, Venetian printer the flood of facts ; it is a mode of being” From: J P Benzecri, “The Soul at the Razor’s Edge” , Les Cahiers de l’Analyse des Donnes, V(2) pp.229-242, 1980 (trans Fionn Murtagh) Richard.Volpato.net

Letting Data Speak -- the Truth Veritas Filia Temporis Truth is the Daughter of Time From: “Statistics is not a means of Francesco Marcolino knowledge, draining into an idea 1536, Venetian printer the flood of facts ; it is a mode of being” From: J P Benzecri, “The Soul at the Razor’s Edge” , Les Cahiers de l’Analyse des Donnes, V(2) pp.229-242, 1980 (trans Fionn Murtagh) Data Analysis now matters as the Material carrier of Culture has changed: • Alphabets • Printing • Media • Data  Today’s Challenge Richard.Volpato.net

Letting Data Speak by: asking questions Look at this data table What do you see? What do you want to know? Smoking Category Staff Row Group None Light Medium Heavy Totals Senior Managers 4 2 3 2 11 Junior Managers 4 3 7 4 18 Senior Employees 25 10 12 4 51 Junior Employees 18 24 33 13 88 Secretaries 10 6 7 2 25 Column Totals 61 45 62 25 193 Originating from Michael Greenacre, now proliferating everywhere! Richard.Volpato.net

Letting Data Speak by: asking questions Look at this data table What do you see? What do you want to know? Smoking Category Staff Row Group None Light Medium Heavy Totals Senior Managers 4 2 3 2 11 Junior Managers 4 3 7 4 18 Senior Employees 25 10 12 4 51 Junior Employees 18 24 33 13 88 Differences that Secretaries 10 6 7 2 25 Column Totals 61 45 62 25 193 might matter Originating from Michael Greenacre, now proliferating everywhere! • Seniority • Authority • Who smokes? Why? Richard.Volpato.net

Letting Data Speak by: asking questions Look at this data table What do you see? What do you want to know? Smoking Category Staff Row Group None Light Medium Heavy Totals Senior Managers 4 2 3 2 11 Junior Managers 4 3 7 4 18 Senior Employees 25 10 12 4 51 Junior Employees 18 24 33 13 88 Differences that Secretaries 10 6 7 2 25 Column Totals 61 45 62 25 193 might matter Originating from Michael Greenacre, now proliferating everywhere! • Seniority Dichotomize all, focus on one interesting value • Authority • Who smokes? Why? Managers Young, Managers Older Folk Smokers Richard.Volpato.net

Letting Data Speak by: asking questions Look at this data table What do you see? What do you want to know? Smoking Category Staff Row Group None Light Medium Heavy Totals Senior Managers 4 2 3 2 11 Junior Managers 4 3 7 4 18 Senior Employees 25 10 12 4 51 Junior Employees 18 24 33 13 88 Differences that Secretaries 10 6 7 2 25 Column Totals 61 45 62 25 193 might matter Originating from Michael Greenacre, now proliferating everywhere! • Seniority Dichotomize all, focus on one interesting value • Authority • Who smokes? Why? Managers Young, Managers Older Folk Smokers Differences (in percentages) can be spoken: See: James A. Davis, “Extending Rosenberg’s Technique for Eg Older folk take to cigarettes more often than Standardizing Percentage Tables” younger people. As they age, some reach Social Forces , Vol. 62, 1984 management where smoking becomes less frequent Richard.Volpato.net ….. Young managers mostly avoid this habit

Letting Data Speak: through visual prompts Macro-views Graphs convert Differences  Distances • Graphs play on memory • Graphs have Grammar • End with macro views • Pacing matters Micro-views Maps show which relationships http://www.worldvaluessurvey.org are likely to be most revealing Richard.Volpato.net

Letting Data Speak by: writing for them Five principles for writing Richard.Volpato.net

Letting Data Speak by: writing for them Five principles for writing • Blind writing – a good graph leaves a clear short term memory – so write with the screen off . Richard.Volpato.net

Letting Data Speak by: writing for them Five principles for writing • Blind writing – a good graph leaves a clear short term memory – so write with the screen off . • E-Prime writing – write without “is”. • Active verbs revivify data! Richard.Volpato.net

Letting Data Speak by: writing for them Five principles for writing • Blind writing – a good graph leaves a clear short term memory – so write with the screen off . • E-Prime writing – write without “is”. • Active verbs revivify data! • Para-graphic ascent : • Encapsulate (start paragraph) • Exemplify (concrete examples) • Explicate (concepts / meanings) • Elaborate (links – causal or conceptual) Richard.Volpato.net

Letting Data Speak by: writing for them Five principles for writing • Blind writing – a good graph leaves a clear short term memory – so write with the screen off . • E-Prime writing – write without “is”. • Active verbs revivify data! • Para-graphic ascent : • Encapsulate (start paragraph) • Exemplify (concrete examples) • Explicate (concepts / meanings) • Elaborate (links – causal or conceptual) • Write along dimensions • Inertia = topicality; • name them using many little contributors • use big variables as supplementary to validate • Follow order of eigenvalues Richard.Volpato.net

Letting Data Speak by: writing for them Five principles for writing • Blind writing – a good graph leaves a clear short term memory – so write with the screen off . • E-Prime writing – write without “is”. • Active verbs revivify data! • Para-graphic ascent : • Encapsulate (start paragraph) • Exemplify (concrete examples) • Explicate (concepts / meanings) • Elaborate (links – causal or conceptual) • Write along dimensions • Inertia = topicality; • name them using many little contributors • use big variables as supplementary to validate • Follow order of eigenvalues • Ecological • Verbs imply activities that interact with other Richard.Volpato.net • Ecology = the map of destiny

Letting Data Speak through: Benzecri’s synthesis Fidelity As Filter (questions) Expertise Experience Informed by Informed by BURT matrix Clustering (concepts) (objects) Data as Disjunctive Matrix

Letting Data Speak through: Benzecri’s synthesis Fidelity As Filter (questions) Expertise Experience Informed by Informed by BURT matrix Clustering (concepts) (objects) Data as Disjunctive Matrix Legacy Flux and Flow of Events Eventualities

Letting Data Speak through: Benzecri’s synthesis Fidelity As Filter (questions) Expertise Experience Informed by Informed by BURT matrix Clustering (concepts) (objects) Data as Disjunctive Matrix Legacy Flux and Flow of Events Eventualities

Letting Data Speak through: Benzecri’s synthesis Fidelity As Filter (questions) Biblical Expertise Experience Informed by Informed by BURT matrix Clustering (concepts) (objects) Data as Disjunctive Matrix Legacy Flux and Flow of Events Eventualities

Letting Data Speak: Ecstatically Bernini’s Ecstasy of St Theresa (based on his own reworking of Veritas Filia Temporis) • The Golden light from heaven - now the light of a questioning quest of high fidelity on earth • The heavy marble floats - now scattered mute data gather to offer insights • St Theresa filled with ecstasy – now the reader can face the future ecstatically • The Logos pierces the heart – now expertise cuts into a problem appropriately contextualised Richard.Volpato.net