Survey Email "Tell us what you think"

[#1]

Quote from: Razer

Matching a general way of writing and word usage/phraseology, spelling/misspelling, punctuation usage/misusage with comments posted on forum and/or in tickets to comments posted in the survey.

Are you sure you're not confusing PlusNet with the NSA?
I would have thought semantic analysis such as this to be beyond the capabilities and will of an ISP.
And, aside from your joke about the heavies, what do you actually think PlusNet would do if they could identify survey respondents?

[#2]

I started to fill it in then realised it was too intrusive why do want exact income and year of birth why not the option of age and income ranges to chose from? Or the option of not answering a particular question?

[#3]

Quote from: CaptainFantastic

And, aside from your joke about the heavies, what do you actually think PlusNet would do if they could identify survey respondents?

I think he's joking

[#4]

Quote from: CaptainFantastic

And, aside from your joke about the heavies, what do you actually think PlusNet would do if they could identify survey respondents?

Send a feeback charge down the customer's line to electrocute them? Is that possible? Assassination could be helpful, sometimes.
My point was, well, to point out it's not necessarily anonymous, but not pertaining to anything serious I made a joke of it. In all seriousness, however, it is within the realms of possibility that being able to identify respondents could influence what the identifier does with that data, seeing as you ask. Not that I'm saying that's what would happen.
EDIT:
Amusingly, just as I posted that, my computer and other equipment on my surge protector shut down. My router, however, didn't. Hmmm ...

[#5]

Quote from: Razer

Quote from: CaptainFantastic And, aside from your joke about the heavies, what do you actually think PlusNet would do if they could identify survey respondents? In all seriousness, however, it is within the realms of possibility that being able to identify respondents could influence what the identifier does with that data, seeing as you ask.

That is, of course, true.
However, my original point still stands: semantic analysis to identify respondents is beyond the capabilities and will of an ISP.
You would need the resources of an NSA or GCHQ.

[#6]

It may or may not be beyond the will of an ISP. You can't flatly say it is. They may at any point in time decide they want to do it for whatever reason. It certainly isn't beyond the possible will of an individual working for them who wants to do a better job, or any number of reasons why they may want or need to do such a thing. You wouldn't need the resources of an NSA or GCHQ.
Come to think of it, though, when it comes to customer surveys, they wouldn't even need to do any such analysis. All they'd have to do is cross reference the logs of survey page calls and sends (time & IP address) with ISP connection logs (time & IP address), oh, and the logs of actual urls called, which they also keep. Instant match to customer ... if they were so inclined.

[#7]

Quote from: Razer

Come to think of it, though, when it comes to customer surveys, they wouldn't even need to do any such analysis. All they'd have to do is cross reference the logs of survey page calls and sends (time & IP address) with ISP connection logs (time & IP address), oh, and the logs of actual urls called, which they also keep. Instant match to customer

I find that more plausible, from a technical angle.

[#8]

I've had some fairly "colourful" survey responses in the past and I've certainly not had the burning desire to find out who they were so I really wouldn't worry too much