Quantitative Information Architecture

Post on 22-Apr-2015

1.714 views 0 download

description

Quantitative methods are applicable for IA thinking be it for hypothesis generation, instrumentation, data collection and analysis of information at scales never before possible with insights that are comparable over time, generalizable and extensible. Quantitative skills can allow IAs to interpret and analyze others’ designs and research more readily, as well as combine methods and models for meta-analysis to help IAs move from description to prediction in designing and developing future interfaces and architectures. This presentation will review why you should use quantitative methods and discuss both foundational and emerging ideas that are applicable for content analysis, behavioral modeling, social media usage, informetrics and other IA-related issues. http://donturn.com or http://twitter.com/donturn

Transcript of Quantitative Information Architecture

Quantitative Information Architecture Don Turnbull, Ph.D.

twitter.com/donturn

#quantia

donturn@gmail.com

http://donturn.com

whois?

•  Software Developer

•  M.S. @ Georgia Tech

•  Software Design & Management

•  Ph.D. @ U Toronto

•  Principal @ Startup (Google)

•  Faculty @ U Texas (Austin)

•  Consultant & Entrepreneur

The fox knows many things, but the hedgehog knows one big thing. Archilochus

•  Quantitative IA is one big thing: a specialty, a mindset •  Designing appropriate experiments •  Leveraging existing quantitative data •  Conducting rigorous analysis

Ways of Knowing & Doing

What does the world look like to a Quantitative

Researcher?

•  What’s Quantitative good for? •  Understanding what users actually do

instead of what they said they do. •  Making comparisons over time •  Generalizable and extensible •  Useful for interpreting and analyzing

others results

Quantitative Methods

•  It is a discipline •  Hypothesis-based •  More applicable to (peer) review

•  It requires a set of skills that have a (much) higher market value

•  It examines characteristics that are constants •  Behavior •  Physical traits & abilities

Why you should go Quant

•  Fight fire with Fire •  Numbers speak the language of business

& technology (C-level execs)

•  (Almost) infallible results •  Qualitative decisions for Quantitative

measurement

The Power of Quant

Predict everything?

•  Anyone can do qualitative research… •  …and anyone does

•  Hard to replicate, hard to validate, easier to do

•  Domain of study (who) is main focus •  Variability is often wide •  Technique is critical

What about Qualitative?

Quantitative & Qualitative should complement each other

Because if they don’t

Things go wrong - fast

•  Computational power & networked systems •  We need new modeling techniques, even

new metaphors to examine the complex systems we interact with

•  Finance, Psychology, Physics & Computer Science

•  Verifiable or provable by means of observation or experiment: empirical laws.

Why Quant, Why Now?

The Network Effect(s)

•  Understanding how people interact •  Metcalfe •  Milgram •  Wellman, Watts & others

A (new) Era of Instrumentation

•  We are undoubtedly in a new era of reasoning

•  Scientific Engineering enabled the original Age of Reason

•  Now to understand intent & interactions

Statistics

Statistics & every day life

•  Weather •  Stock Market

•  Whole channels on TV devoted to both

•  Sports…. All the time…. Everywhere •  Your Net Worth, IQ, Postal Code, SAT

score, GPA...

Data Science

A New Kind of (Empirical) Science

What’s next ?

Analytics

HCIR (IRUX)

Atomic Information Architecture (a marketplace)

Pervasive, Emotive & Suggestive

Summary

New Age of (Empirical) Reason

Quant at scale = new Qual insights

Data Science (& its affect on IA/UX)

What’s next isn’t that far ahead…

Thanks

donturn@gmail.com

http://donturn.com

twitter.com/donturn