Categories
Opinion Quinnipiac Social Media Social Media Class

Quinnipiac Assignment 12 – ICM 552 – Privacy and Big Data

Quinnipiac Assignment 12 – ICM 552 – Privacy and Big Data

The Price of Handing Over an Email Address

Quinnipiac Assignment 12 – ICM 552 - Privacy and Big Data
The old MSN Hotmail inbox (Photo credit: Wikipedia)

One of the best pieces of advice I ever got when I was first surfing the Internet and beginning to understand the online community was to get a private throwaway email address. The idea was to use an online provider (I originally used Hotmail, and then moved over to Yahoo!) and not give out the address my husband and I had gotten when we signed up for our Internet Service Provider, Brigadoon. Brigadoon is long gone, replaced by several iterations and that service is now provided, in my home, by Comcast.

Eighteen years later, the Yahoo! account is one of my primary email addresses. Although my husband still uses the Comcast address, I almost never do.

It was an odd thing, back then, to use a separate address. We didn’t do this offline, e. g. neither of us had a post office box. Was it an unreasonable push for privacy in a marriage where we had vowed to be open with each other? Or was it a reasonable need for a separate space, almost like a separate set of friends or a man cave?

Of course, as we began to be spammed, I learned why this was such a good idea.

Throwaway Email Addresses

In fact, I also learned that using Gmail was better for activities such as job seeking. Now my resume sports a Gmail address, even though I still read most of my email via Yahoo!

But the throwaway address itself has become a more predominant one for me. And so now I am finding I don’t like it quite so much when it’s put out there.

LinkedIn and My Email Addresses

Once again, LinkedIn is a bit of a bull in a china shop when it comes to email addresses. I currently have several addresses on my account, some of which are no longer active. However, if I attempt to apply for a job through the LinkedIn site, a drop down menu appears where my email address will be added. There is no opting out. You have to pick an email address to be sent along with your application, even though it’s possible to communicate on LinkedIn itself. Your telephone number can be altered or deleted, but not an email address. You have to send one along to whoever posted the job.

Does this compromise privacy? I think it does, as there are a lot of reasons why I might want to remain a bit hidden when applying for a job. Employers are able to post jobs anonymously, but potential employees aren’t being given that luxury when it comes to applying for those same openings. It’s just another example of potential employees and their possible future employers not being on anywhere near the same level.

And maybe, just maybe, LinkedIn should rethink this policy, and give job seekers an opportunity to hide themselves better, at least when the initial application goes out.

Categories
Analytics Quinnipiac

Quinnipiac Assignment #02 – Qualitative and Quantitative Analytics

Quinnipiac Assignment #02 – Qualitative and Quantitative Analytics

This week’s assignments at Quinnipiac were centered around the differences between quantitative and qualitative analytics. I had a couple of essays to write.

I have decided to reprint one of my essays here, in its entirety.


Qualitative versus Quantitative

I think that a choice between the two is, perhaps, misplaced.

English: Interactive Visualization of Qualitat...
English: Interactive Visualization of Qualitative and Quantitative data in a web based mixed methods application (Photo credit: Wikipedia)

Aren’t both of these necessary, in order to really see the big picture?

We Love Quantitative Data

Probably the best part of quantitative data is that it’s relatively easy to obtain, particularly online. Consider this – do we, given the current state of technology, know everyone who comes into, say, a department store?

Web analytics framework
Web analytics framework (Photo credit: Beantin webbkommunikation)

Even when we break this down to hourly increments, and even if we look at closed-circuit cameras, we still might miss someone. After all, if a person leaves and comes back later, we might not notice that it’s the same person.

Not so with the web. Cookies and other tracking codes give us the ability to know that a device has returned; and an account if our site allows for user accounts. That still doesn’t help us if everyone in a household uses the same account, but it’s a start.

We look at our web data and we think – aha! User #12345 has returned four times in one day!

And then we have no idea why that happened, and no way to capitalize on it. It’s the ultimate in vanity metrics, e. g. it’s stuff that can be measured but it isn’t necessarily actionable, or even desirable information.

We Love Qualitative Data

With qualitative data, we get more into the whys and wherefores.

coding cat-egories
coding cat-egories (Photo credit: urbanmkr)

Why did User #12345 return four times in one day? If a purchase is made on the fourth go-‘round, that’s terrific. But why were there three other visits? Even someone performing research and then returning later might not necessarily visit two more times. What’s up with that?

Maybe the website was slow those two other times. Maybe User #12345 got busy and abandoned the cart for Visit #2 and Visit #3. Some of this is inferential. Some of it can be proven, such as site slowness or at least traffic spikes that could imply speed issues. We can’t get into User #12345’s head (at least, not yet).

We REALLY Love Them Together

I think we’ve got to look at the two types together.

In the Huffington Post article, The Big Data Craze Is Just as Qualitative as It Is Quantitative?, Sean Donahue writes, “But for brands, political campaigns and advocacy organizations that aim to have data-driven conversations with audiences, it will be more important than ever to apply qualitative logic and human reasoning to online analytical models. In short, subject matter expertise and deep knowledge will matter more than ever before given the rise of big data.

As communicators, even with what we have at our fingertips today, we need to immerse ourselves in the substance that contextualizes big data and allows us to make sense out of it. This means committing more time, asking more questions, consuming more content and never losing sight of the fact that data without actionable insights is meaningless.”

I believe that what Donahue is saying is that we can and will be getting great big garbage bags full of data, and soon even more of it will be at low or no cost. But without contextual analysis, it’s somewhat meaningless.

Further to that is Anmol Rajpurohit’s point in Qualitative Analytics: Why numbers do not tell the complete story?, wherein he writes, “Quantitative analytics still needs more manual intervention and the results are often fuzzy. In absence of a clear-cut approach and thus automation, it is not as time and energy efficient as the traditional quantitative analytics. But, qualitative analytics is still indispensable as it provides deep, actionable insights about the ‘why’ and ‘how’ aspect, which often gets ignored as we continue to be inundated with the ‘what’ ‘where’ and ‘when’ of statistics.”

As Rajpurohit indicates, qualitative data is fuzzy and manual and not automated. It’s a slow process (and perhaps a less exact science than quantitative), yet it remains necessary to a holistic understanding of online data.

06/30: It's Peanut Butter Jelly Time!!!
06/30: It’s Peanut Butter Jelly Time!!! (Photo credit: ttyS0)

To me, this is peanut butter and jelly. They’re fine separate, but they work best together. With purely quantitative data, we can know that a particular literary passage has 278 words. We can know that two of its longest words are twelve letters long: consummation and undiscovered. We can find that the (probably) most frequently-used word is ‘the’, with twenty occurrences. With qualitative analysis, we learn that the mystery passage is Hamlet’s To Be or Not to Be soliloquy. With qualitative analysis, it stops being a laundry list of words, and its context affords a meaning that goes beyond bare statistics.

References


Enhanced by Zemanta