Another way metadata is being utilized: Facebook keeps details of what users type in their status but may never publish.
So I don’t recall us bringing up this particular example when we talked about analytics and ways it can be used that no one imaged, but stumbled across this study today so thought I’d share as I consider it pretty clever — particularly because racism is such a hard topic given many people don’t necessarily answer honestly. Harvard economics doctorate candidate Seth Stephens-Davidowitz paired the frequency of Google search words with racist tones and mapped it against how Barack Obama performed in his first election. The link included above is just one write up of the study. A Google search here will show you many more, including links to the original paper.
Without a doubt, there is a need for additional privacy standards when it comes to big data, not just to guide industry but also to help individuals understand the scope of what they are at risk of divulging when they so easily give information about themselves away. As Designing the Personal Data Stream: Enabling Participatory Privacy in Mobile Personal Sensing pointed out, the Codes for Information Privacy are just inadequate for today’s mobile technology and the big data it generates. The authors’ proposition, that individuals need to be given the power to control which of their big data streams are collected and analyzed by using an intermediary “vault” is quite appealing to me. It’s a bit like having curtains on a window, depending who is looking in, I can decide how much to show them by drawing back the curtain a little or a lot. I wonder how balky such a system would be to use initially, but could see great marketability for such a system.
I was a user of instagram. (Picture to left is one I posted at the Democratic National Convention last year when I was covering it…no idea if it was Michelle Obama, Barack Obama or Bill Clinton…). But late last year when the company’s announcement highlighted how my pictures might get used, I quickly stopped using the site for anything personal and am more cautious as well when it comes to Facebook. Still worth it for professional reasons, but on personal note, the price for these free services is privacy, and that’s a steep one. Personal Data Stream architecture is appealing. It is the first idea I’ve heard about in a long time that gives me some hope that there may be a way to restore some privacy to this system. I might be fine with letting OnStar know, for example, how much I drive, but I don’t necessarily want it to also be collecting my GPS coordinates. Sure, they may tell me they won’t store the GPS data, but can I trust them? Giving the power back to the individual to monitor the data collection is more inline with America’s right to privacy principle.
Finally, the story about medical data is not unique in the sense of there has always been tussles over who “owns” medical records — hence the recent fight in Tallahassee over how much doctors’ offices and/or medical record firms could charge you to get copies of your own medical files. But what is shocking now when it comes to medical devices is the quantity and quality of the data. Clearly, there needs to be new standards and/or laws pertaining to devices implanted in the body.
I posit one other thought about medical devices. It’s not just access to the data but it’s also that the individual with the most to lose is not in control of who else has access to the data. That’s what former Vice President Dick Cheney’s doctor worried about — that in the wrong hands, information about the former vice president’s heart defibrillator and heart pump could be easily hijacked to cause a heart attack. And you may have seen the Homeland episode with the same premise.
1) How have privacy concerns impacted your use of social media? Are there things you don’t do as a result? How diligent are you in monitoring privacy settings?
2) On your smart phone, do you restrict applications from accessing your data? Why?
3) My employer has made it clear that what I do on social media is an extension of my professional life and I am expected to adhere to the same standards as if I was working — but never has asked for my password. What kind of expectations are there in your job about using social media?
So maybe we’ve all been going about this wrong. This isn’t about how the Internet and social media are reinventing marketing — it’s about how the past 200 years didn’t rely on social networks. That’s according to a review of the new book Writing on the Wall: Social Media—The First 2,000 Years that says “modern mass media was a 200-year aberration: a function of the high production costs associated with large printing presses and broadcast machinery. The rise of the Internet made everyone a potential publisher, and, thus, media reverted to its natural, social state. Everything old is new again.”
The review in Washington Monthly is worth a read.
So the first thought I kept having during this week’s reading? We all should be information specialists. There is just more data than there people capable of organizing and shifting through it. The bad rap a librarian’s masters got a few years ago? I just don’t buy it, particularly after reading this article about the vast opportunities and future marketing potential. Pair an information science degree with some computer skills and bet you could write your way into a marketing firm or major Fortune 500 company. Everyone has this data, but they need mechanisms for accessing it in a timely way.
The second thought was that Facebook’s founder was brilliant. This article puts into great context the efforts Facebook is making in organizing data for third-party use. Mark Zuckerman is going to one day rule the world (if he kinda isn’t already).
Thirdly: The need for some kind of governing mechanism for this data collection is vital. But can one actually work? The article mentions the Digital Advertising Alliance’s Self-Regulatory Principles for Multi-Site Data. But without some stick, some data collectors would never comply.
My fourth thought: We don’t even yet know what this looks like. Consider, for example, the entrepreneurs featured in Monday’s New York Times who are collecting odd data but finding big customers for it.
Finally, count me among those privacy advocates worried a bit about the incredible potential for abuse. Yes, do I love having only relevant ads pop up on my web pages when I’m surfing the Net? Of course. Don’t waste my time. I’m also an advocate for democracy and while the Internet has done a lot to give voice to minorities who otherwise aren’t heard, I also worry that the Internet has allowed more fragmentation. It’s no longer that there aren’t bowling leagues, but that individuals never really have to break out of their comfort zone. Don’t want to ever hear bad news? Not unlike Aldous Huxley’s Brave New World, data utilization may soon make that a reality.
1. A hypothetical case study: Money and technology are no object for your client who sales children’s jewelry, so what kind of data could you envision mining to optimize a micromarketing effort?
2. Do you still fill out customer surveys that seek input on your household income, education level, etc? If so, when was the last time you encountered one?
3. Are secrets passé? If there was information in your personal life that you did not want to share with anyone, do you think there is any mechanism now to do so assuming a fragment of it may have been documented by a technological device?
So anyone tried Lightbeam, a new cookie-tracking tool from Mozilla’s Firefox, yet?
I was reminded this week how LONG it has been since I had a statistics class…and of the difference between scientific and applied research. Building on what Richard Kalehoff already noted in his post this week, I was struck that this kind of intensive research is very different from what I imagine would be utilized in most professional settings where you could tap the power of online polling, surveys, A/B testing, etc. to quickly draft a plan specific to your client’s goals.
All that said, I do find the kind of indepth attitudinal research the Philadelphia campaign undertook very interesting and appreciate that it provides guidance for other campaigns on children’s nutrition. In the pre-post survey, I was struck about how different message delivery systems (TV, display, radio) had different impacts and that overall, the motivators most successful were factually (weight gain and diabetes prevention) not emotion (doing something good for your family), except in the television commercial. That makes sense to me in that in the TV commercial parents might connect more with the mother than they would a static image in print or audio-only in radio.
Still the research found that if a parent connected on the “doing something good for your family” they were more likely to have a higher intention of reducing the beverage consumption than those who reacted to “weight gain.” I wondered if that was personal baggage – the parent might be obese as well — so I poked around and found this article about the study and found Amy Jordan, one of it’s authors, explain:
” First, they (parents) don’t notice when their kids gain weight. ‘We found that although parents were able to recognize when they themselves were overweight, they didn’t recognize when their children are overweight,’ she says. Although as many as half of the city’s children are overweight, only 20 percent of parents notice. That may be because as more children in the neighborhood get heavier, it’s harder to notice their own children’s weight gain.”
Canadian vs. Chineses web behavior
Reading the online behavior research comparing Canadian college students and Chinese students in Canada, I found myself really wanting to see which websites they used for testing to get a better sense of what they classified as low task relevance vs. high task relevance. But in the end, I wasn’t surprised by the bottom line: “Service providers should enhance the hedonic aspects of the website for the Chinese and the utilitarian aspects for Canadians.”
1. So let’s pretend Amazon.com was one of the sites used in the Canada/Chinese study. What would you classify as a low task relevance vs. high task relevance feature?
2. Do you think the findings in the soft drink study could be translated to other campaigns about children’s nutrition, say such as eating more vegetables or less sugar over all? Why or why not?
Wow. Was just wrapping up presentation for Wednesday night when I saw a reference to the New York Times amazing online piece from Sunday that seems to invisibly adhere to all the findings of eyetracking research in an unprecedented way for traditional-style journalism. I encourage you to take a look.
Poynter weighs in with a writeup.
From the time I first became aware of eyetracking technology, I have always thought of the Borg, that cybernetic alien race in the Star Trek universe that tries to assimilate other species into its collective – by whatever violent means necessary (“Resistance is futile!”). Of course it was a visual association with the contraptions that eyetracking research has required in its most common form in recent years (research subjects don eyeglasses that have cameras mounted on them). But after reading these four articles (and doing some other research for my presentation on Wednesday), I wonder if the association isn’t a little deeper.
Because what is eyetracking trying to discover? How we see and what we see with the idea that if that can be ascertained, researchers have a better idea of how to communicate with us. Or, in some cases, even understand us, apparently, as the introduction to eyetracking article made a fascinating mention of how scientists are using the technology to gain a window into the brain. Was it Cicero who said the eye is the window into the soul?
I was already pretty familiar with content in the two Poynter articles due to my job in the news business: The 2007 Poynter study comparing print and online presentations of newspaper content and the 2012 study on news content viewed on a tablet. In the context of this class, I find them a great case study of applied research for a particular industry. The 2007 study feels very much like a product of its era, as newspapers were trying to understand web readers as extensions of print readers rather than wholly separate products – and they had not yet seen the dramatic drop in display advertising revenue for their print products. Six years later, that has changed dramatically. The tablet study in 2012 reflects very much the latest evolution of trying to understand the platform for delivering a product.
The Google map white paper felt a bit more generic in its applicability, and informative for all industries trying to leverage SEO efforts. The upshot: If you aren’t ranking at the top of Google+ map searches, get yourself some social media links or reviews because they make a difference in eye traffic.
1. Does the Google map research jive with your own personal use of such search functions? Were those heat maps your heat maps?
2. The 2007 Poynter study found people, once they commit to a news story, actually read further online than in print. I remember when those findings came out and I assumed that was at least in part due to a click of the mouse is easier than refolding a newspaper to find the jump. But do you think the findings are still even true six years later? Or are we all just serial scanners? When do you spend a lot of time on text?
3. Before this week’s readings, were you familiar with eyetracking? If so, in what context?
Stumbled across this in my Facebook feed this morning from someone who had to strand a Hertz rental car in the Colorado flash floods and was none too happy with how the company has handled it:
“…You’d think when the President declares a natural disaster, Hertz management might take some measures to address it, let alone not embarrass themselves. And the thing is—I’m almost positive this is all covered by insurance. Help me help you, Hertz. Cut the Kafkaian Mobius Strip bullshit and help your customers work things out, so we can all move on to the important things, like you screwing the insurance company, and me getting screwed by it. “
Read the full blog post here. I’ve bookmarked the page in case we find out Hertz responds.