hockendougal

Thanks for the that Moshu, we’re not trying to screw anyone here (like really).
1. That wasn’t really the goal here, and based on our previous experiment about half of clicks to the theme download also click to the license page (*see below, it is not as prominent as it should be) – although they may not be really reading it. I hope that everyone who installs the theme understands exactly what is going. Our goal is not to ‘trick’ people into giving us data. We got plenty of data the last time from people who seemed excited and interested in the experiment.

2. You’re right, I’ll have this adjusted.

In response to 3,4,5 – I hadn’t realized the the license link was not there, I have personally edited this right now to make it prominent (above the fold, at least on my screen) and I’ll make sure it gets cleaned up better later. The reason for that “we discourage anyone to use it” is because while we are relaxed people, not everyone is. The legal speak for the license is still under review. The ‘human contract’ is there, and that is essentially what you are agreeing to. Will it stand up in court? Probably not, nor would we ever come to a situation where we expect it to. There should probably not be the expectation that you are agreeing to something that you cannot see, so I will have the wording changed accordingly – when the legal code is approved, it will be on a new version of the license (so no retroactive agreement)

6. I really hope not. Like I said, that’s not the goal.

For your information, the details are the previous experiment are in the posts here: https://www.creativesynthesis.net/blog/?s=recycled+canvas

Specifically this and this provide some overview / results information.

One other big change we’ve made is that previously the code wasn’t directly integrated with the theme (it was in a separate plugin which we required to be installed, but which could be easily be deactivated). Is that semantic separation of tracking important enough?

Huh well, this is interesting because our previous experiment met with almost no criticism (which I thought was surprising at the time). I’m unsure why the reaction is negative this time and it wasn’t before. Perhaps I should make a few things clear:

* We aren’t a commercial entity or ‘professional’ theme group. We’re an academic (nonprofit) research group who finds doing research in blogging and in wordpress (specifically) valuable and fascinating. Our last experiment was a ridiculously positive success and I was personally hoping this would be too. The experiment is more important then the theme for us (the theme is only a carrier of the experiment).

(I should add to the above that this means we have both a set of ethics and oversight on data collection)

* I’m happy to have someone make changes based on feedback to the data collection, but there will always be a little bit of data collection. I hope that people don’t just strip the theme, there are plenty of wordpress themes. We usually encourage people who use our theme to think of themselves as ‘junior experimenters’ helping gather data for analysis.

Edit- It’s really important for us to go about these things in a way that is acceptable to the wordpress community, while acknowledging that not everyone will love it. Any feedback on this would be really appreciated.

Jeremy, if the email address is the really obnoxious part we can remove it, and come up with an alternative mechanism for identifying the data owner.

I completely agree with you in the general sense. The reason we offer the paid / free with experiment versioning is so we can get a good mix of people who don’t mind being in the study while still allowing people to opt out and just get the thing if the experiment is too obnoxious for them (while still supporting our research in the monetary sense).

We haven’t completely finished the backend for this, but one of the reasons for collecting the admin email is to allow for owner access to their data. As to the rest of the data, its anonymized on our side – we only see a hash when we look at the data, and any textual content (drawn from rss) is run through a semantic parser so we are left with no real sense of individual experimenters (and this is important when the data is made public, which the license requires that it is)

The previous experiment was slightly different in its mechanism. The experimenter stored the data locally and only pinged us with the location of the data, which we harvested occasionally. I’m curious if you (or anyone else) think there are better ways to go about this / things we can do to make people feel better about the process. Historically when the experiment is over we’ll release anything left under the gpl, but that will probably be a year’s time from now.

I agree that the information is fairly invasive, but we do spell out exactly what we’re tracking (there is no mystery here). As to the why, we’re doing a study on theme usage and distribution statistics.

We did a similar thing before with the recycled canvas theme which was probably more invasive (tracking mouse patterns of every single user on your blog). You’re under no obligation to use the theme, or you’re welcome to buy the other version.