Reality and Perception of Copyright Terms of Service for Online Content Creation (CSCW 2016)
Do you understand the terms that you “click to agree” to when you post content on sites like Facebook, Craigslist, or DeviantART? It’s okay, neither does anyone else! But whether blog posts, photographs, or social media status updates, you own the copyright in your original content – which means that in order for one of those websites to be able to display it, you have to license them the right to use it. But what are those licensing terms, and do people understand what rights they are handing over in their work? For example, here’s the license that all YouTubers provide in their videos:
worldwide, non- exclusive, royalty-free, sublicenseable and transferable license to use, reproduce, distribute, prepare derivative works of, display, and perform the Content in connection with the Service and YouTube’s (and its successors’ and affiliates’) business, including without limitation for promoting and redistributing part or all of the Service (and derivative works thereof) in any media formats and through any media channels
If you understand exactly what that means, you’re one of a rare few! And if you want to hear about a license that’s even more far-reaching, you can read my blog post about what LinkedIn’s copyright terms used to be.
My initial work on this topic was published as a CHI WIP in 2014, and was featured in The New York Times. Encouraged by some interesting results about the content of copyright licenses in Terms of Service (TOS), we (myself, my PhD advisor Amy Bruckman at Georgia Tech, and Cliff Lampe at University of Michigan) continued the research, using a survey to match user perceptions of copyright terms to the reality that we gathered from content analysis of the terms themselves. The results of this work are upcoming for CSCW 2016, and there is a lot more detail in the paper, but here are some highlights:
- For the 30 user-generated content sites we analyzed, the average word count was 3,852 words, with an average grade level reading score of 14.8–that of a college sophomore. At an average adult reading speed, it would take almost 8 hours to read the TOS of just those 30 sites.
- Only five of these sites included any kind of readable, plain-language explanation of their terms.
- There was a large amount of variability in what licensing terms were on the sites–in other words, it’s not all boilerplate. Just because you know what the terms for YouTube are doesn’t mean that you know what they are for Facebook or LinkedIn.
- If we lump websites into categories (e.g., social network sites or music sites), people favor social network sites (like Facebook or TwitteR)
- How well people can correctly identify which terms exist for which type varies widely based on the term – there are some that would be consistently surprising to people. For example, you provide most sites the right to “modify” your work, but survey participants typically thought that this was not the case.
- Some of the terms that people don’t like (that they thought shouldn’t be there) are also the most surprising. For example, an “irrevocable” license appears more frequently than people expect–and if they knew that, they probably wouldn’t be happy about it. In other words, if you ask someone if a website should have a license that the user can’t revoke, they typically say “no,” but also “that’s okay, because I don’t think they do that!” And in many cases, they would be wrong about that.
So what does all of this mean? The short version is that there is an obvious usability problem here. There is information that users cannot understand, and it is (based on our findings) information that they would care about if they were aware. Considering how much HCI research goes into so many aspects of design, how often do we see human-centered TOS design? In this paper, we call for more attention to this problem. Even when websites are bound by lawyers to include certain language in TOS, including plain language explanations – not only of what terms mean but what websites mean to do – would go a long way.
Also, this study (conducted a year and a half ago) was my first time recruiting research participants from Mechanical Turk. It was an interesting experience, and since then I’ve thought a lot about the process, and spent some time as a turker myself in order to better understand the platform and the worker experience. One side finding of this work is that very few participants failed our attention check – only 14 out of 410. Based on our data, we felt very good about the quality of responses that we received from turkers.
For more detail, see the full paper. I will be presenting this work at the conference on Computer-Supported Cooperative Work & Social Computing (CSCW) in San Francisco in February 2016.
Fiesler, C., Lampe, C., and Bruckman, A.S. Reality and Perception of Copyright Terms of Service for Online Creative Communities. CSCW ’16: Proceedings of the 19th ACM Conference on Computer Supported Cooperative Work & Social Computing, (2016).