Last week, I asked a diverse group of researchers to share their thoughts on the future of measurement. The researchers (names below) loosely have 1-3 common denominators across psychology, artificial intelligence, text analysis, social media, and/or advertising research and analytics. As witnessed in the visualization of our discussion (courtesy of Wordle), data was the currency of conversation. Broadly, responses spanned tools, methods, metrics, constructs, and value.I categorized responses into 4 major predictions about what 2009 will hold measurement-wise. In the pdf, available for download below, I've extracted verbatim from all respondents to support the predictions. The discussion took the form of a Google Group, so in some cases, it was necessary to remove substantial dynamics of conversation. I’ve done my best to capture the essence of themes represented and hope this aggregation will be insightful in understanding where things have been, and potentially where they are going.
I walk away with a sense that researchers are optimistic, yet there’s an undercurrent of “let’s get real.” What are we really measuring—what is a meaningful connection, what do we really pay attention to...
In other words, to paraphrase Diane Court (Ione Skye) in Say Anything:
I’ve glimpsed our future, and all I can say is: Go Back.Let’s get back to basics in 2009 and measure something meaningful.
Predictions:
- We will substantially advance our understanding of individuals and the meaningful connections they have.
- We will identify methods to tap what people are *really* thinking, feeling, and paying attention to, meanwhile gaining insight on what a measurement is truly capturing.
- We will determine how to measure the value of social interactions and attach financial value, whether we’re monetizing attention or a new medium.
- We will build better tools to manage-- analyze and visualize-- massive volumes of data, primarily tapping the evolving social graph.
Thanks to the many researchers who contributed:
Neal Burns, University of Texas, Austin Center for Brand Research
Walter Carl, Chat Threads,
Maury Giles, GSD&M Idea City
Sam Gosling, University of Texas, Austin: Department of Psychology
Seth Grimes, Alta Plana
Matthew Hurst, Microsoft Live Labs
Paul Janowitz, Sentient Services
Shawn Kung, Aster Data
Roddy Lindsay, Facebook
Jamie Pennebaker, University of Texas, Austin Department of Psychology
Martha Russell, Media X at Stanford University
Miles Sims, Small World Labs
Marc Smith, Telligent
Daniel Tunkelang, Endeca
10 comments:
Kate, The future of research?
1. We'll increasingly wake up to the fact that what people say is not really what they mean. And that people who do offer to say what they think they mean are nothing more than anamolies. Hence, surveys will lose even more clout.
2. Real, UNPROMPTED observable evidence will become more the norm.
3. Trusted triangulation of different evidence sources will become core of insights and solutions to important questions.
Finally, one interesting futures question about research is how findings or interpretation spread and influence opinion. There's often a huge disconnect between core, technical findings, and the digestible, mutated bytes that tend to become the perceived spirit of the findings as they disperse. Having spent most of my career building the reputations of emerging research and analytics companies, I can say the context in which this happens has changed drastically over the past 15 years.
Kate - |get a page not found when I click the link. Is it me or is the link dead now?
Best
Haydn
Thanks for your thoughts, Max - Your first 2 points are very consistent with the responses categorized under Prediction 2. The gist was completely anti-self report-- about identifying ways to capture nuances, feelings, attitudes, etc. through things like linguistic style.
Triangulation will be important, as will discriminant validity-- see an insightful comment made by Sam Gosling, also categorized in #2.
Lastly, I agree that's a fascinating area and one we'll see much growth in via social technology. Hopefully facilitating discussions like this one and encouraging transparency wrt what people are working on will help researchers disseminate their findings in a more accessible way-- bypassing the mutated byte or availing the source...
Hmm, not sure why that's happening. You should be able to both view and download here: http://www.scribd.com/doc/9405392/Social-Abacus-Future-of-Measurement-2009
Email me: k.niederhoffer@gmail.com and I will send via email if the problem persists.
thanks to you, Daniel, for kicking things off and staying involved!
Great summary from a lot of the folks doing the best thinking out there.
At Crimson Hexagon we're focused on solving the problems around understanding non-solicited opinion and managing/analyzing massive volumes of data. There's also a balance to be struck between identifying/tracking individuals, particularly influencers, and being able to distill the large trends. Thanks for the insights, Perry
Thank you, Perry. Massive amounts of data with naturalistic opinions is a rich source of answers. [I too used to work in brand monitoring]. I'm curious as to how you operationalize influence. Do you think influence is agent-based-- as opposed to more 'unit'/ meme-based, or something more of a connection?
Thank you for the article. I enjoyed reading it a lot.
P.S. Scribd requires registration in order to download the article; it would be nice if it was possible to simply download it from your website.
We're still looking at different ways to crack the influence nut -- in answer to your question, I see it as a bit of both and don't think we (or anyone, yet, that I've seen) have arrived at the Right Answer. We're doing some independent calculations, but are also working with partners (as well as specific clients) to define what influence/reach mean. Some of this will be standardized from specialist firms, and I look forward to seeing those efforts.
Cheers, and happy New Year,
Perry
Unfortunately that's a drawback of the blogger platform. I also uploaded it to a google site, but the advantage of Scribd is the ability to view and/or download. Try: http://sites.google.com/site/futureofmeasurement/
Post a Comment