 Kia ora koutou. My name is Michael Descardes and I'm the user experience manager for the National Library of New Zealand. One of the services managed by the team of which I am part is Digital New Zealand, Aotehe o Aotearoa. It's a search engine and data service that gives access to 27 million cultural heritage items for more than 200 content partners, all in one easy to use place. And today I'm going to be presenting a bit of work that we've done around rights on the Digital New Zealand service. For those who would like a copy of slides or want to follow any of the links, there's links and annotations in the slides. The little digital program that comes with NDF, you can find it linked there where that arrow is. So Digital New Zealand, what do we do? Our mission is to make New Zealand's cultural heritage easy to find, share and use. Historically we've done pretty well with the find part as our tools for keyword searching, faceting and filtering have been strong. Our tools for sharing have improved as well with the addition in recent years of Digital New Zealand sets and stories which let people easily collect and share the items that they find. However, we weren't always doing the best job that we could at letting people know how they could use the items they find through Digital New Zealand. We saw this as an opportunity for improvement and we set out to see how we could make it better. The humble right statement is at the heart of things when our institutions patrons come into closest contact with the items from our collections. Without a quick and clear understanding of rights in context, we run the risk of a patron using an item in a way they shouldn't or I think more tragically believing that they aren't allowed to use an item when in fact that they are. But determining copyright is complicated. This is an excerpt from just one page of Victoria Leachman's 12 page flow chart about determining copyright at Tapapa. We cannot reasonably expect our poor researchers to navigate this logic on their own. Our job as Information Access professionals is to distill this flow chart into an answer to one question. How can I use this thing? At our library this is an important problem to solve and in fact in 2013 when an internal working group was charged with building a five-year roadmap for our online efforts, we rated clear and consistent right statements as the very first issue to tackle. And we're not the only ones struggling with the usability of copyright information. The rightstatements.org coalition led by our counterparts at Europe Piana and the Digital Public Library of America is making an attempt at coming up with an international standard of right statements. A bit more about them later. Up until 2017 an item record page on the Digital New Zealand.org website looked like this. An about copyright page elsewhere on the site offered an overview of our approach to copyright. But in the context of a specific item that you were viewing, we didn't offer much help. Pretty much the only information about rights and usage information that we displayed came from a single field called usage. This is a controlled field that we use to allow the filtering of searching, the filtering of searches by how you can use an item. The usage field has five possible values shown here and it can have more than one value at once. While this is no doubt useful, the value by itself, sometimes as little as a single word, offers scant guidance for the uncertain user. Additionally, the choice and placement of a question mark icon for more information link made us look like we weren't quite sure about the little we've offered. A click on that icon offers a bit more commentary, but the gist of the statement was more or less, we're not sure. Check with the rights holder. While this statement had the virtue of being true and tried to be helpful, a bit of user testing led us to realize that it's full of mild ambiguities and potential misunderstandings for the casual user. Some of this ambiguity stems from the fact that digital New Zealand is an aggregator of other people's content. Unlike many other museum or library websites, we are not the rights holder for the vast majority of the items to which we link. So we need to walk the line between making solid statements about use while making it clear that the ultimate call about the usage rights rests with the rights holder, not with us. When we refreshed the design of digital New Zealand.org in 2017, our team had an opportunity to make the item record pages better and to move towards the high priority goal of clear and consistent rights statements. We started by building rough prototypes of the kind of thing we wanted in order to get them in front of real people as soon as possible. Our first draft wasn't that earth shaking, and in fact it offered pretty much the same monosyllabic statements as the old site. We did, however, swap the doubtful question mark for a slightly more confident exclamation point. We also tightened up the language in the pop-up for a bit of clarity, but this was still pretty weak sauce in terms of user experience. For the second iteration, we set out to dedicate a significant chunk of the screen real estate on each item page to explaining rights and usage status in detail. So this was the first version of the new rights experience. Our goal was to see if we could lead with a single question, can I use this and answer it as well as we can and then offer contextual detail. At one point in the process, we were even concerned about making two definitive statements about the rights status of items on behalf of our content partners. So we even tried this cheeky statement to see if it would grab people's attention and force them to read the fine print more closely. We actually prototyped that. So the primary message that we delivered in our first version of this traffic light for rights looked like this. The question, can I use this? Had four possible answers loosely matching the conditions on the left. With our first prototype in hand, we set out to get some feedback from as many real live human beings as we could. We included the new rights treatment in approximately 20 one-on-one user testing sessions being conducted for the wider site redesign. We posted the prototypes on a private server and invited our content partners to have a go and offer their thoughts. We also had some sit-down review sessions with a few friendly experts. I'd like to particularly thank Victoria Leachman, Sarah Powell, Adrian Kingston, Mike Dickison, some folks who were in the room, and Mandy Hank for checking our work and giving great feedback at various stages. And if I've left anyone out, my apologies. But we did have some really good feedback from folks. The first major thing we learned from the feedback was that people like the question and answer format, but that the answer to our initial yes-no question, can I use this without permission, was too often yes, but or possibly with conditions. So this was our version two approach, which presented the copyright status a bit more traditionally up top. And then we went into detail with answers to the question, what can I do with this item? We also added some more specific language whenever we encountered a Creative Commons license. Creative Commons license URLs are very specific and machine parsable, so it was relatively easy to write a bit of script that presented the very specific information for the appropriate license. This bit of language here was also important for our role as an aggregator, where we essentially say in the friendliest way possible, this is our best information right now about the rights, but here's a link to the original item on the site of the rights holder, and it's your responsibility to make sure that you check with them before you do anything drastic. There's also a section about what you are allowed to do with any item, regardless of rights status, but I'll go into a bit more detail on that a bit later in the talk. We spent a lot of time getting the logic that controls the display of the statements correct. The inputs of the logic are three fields from the Digital New Zealand API, which is usage, rights, and rights URL. The data that populates these fields comes from our data harvest, and the initial logic for harvesting that data varies by content partner and collection. For our purposes, the key fact is that usage is a required field and must have at least one of these five values, so that was something that we could say would always be there to be depended on. For those that are interested under the hood, this is the final version, somewhat simplified, but this is most of the logic, kind of start in the upper left, you go through, if usage is all rights reserved, we go right into the all rights reserved bucket, then we switch and say, is the rights URL a Creative Commons license? If so, parse that. Then we look at usage again, is it unknown? That gets you the gray icons, and after that we go into default, we look for share and modify and specific bits. Then we swing around, does a rights URL exist? If so, we add a little bit that says, here's some more information that the rights holder provided, and we send people off there with a link, and the rights field is essentially a text field that has more rights information that we might have harvested with special usage, and then we'll display that text under a little more information icon. So once we had the logic sorted and tested, we were ready to roll our shiny new rights statements out to the digitalinset.org website. As soon as we had a product that made us, our content partners, and our lawyers happy, we deployed the rights block to the new item pages. This is what it looked like when it went live earlier this year, and it still mostly looks like today. Where previously the rights information was conveyed with just a word or two, we now dedicate about a quarter of the page real estate to communicating the rights status and the rules for reused in the plainest possible language. We had our designer do a pass on the graphic design and typography of the rights statement block, including these snazzy new icons, ensuring that each statement was easy to read. The mobile view uses the same iconography, but collapses the additional commentary. So far, the new statements have been very positively received by our users. We had an improvement in our satisfaction score from this year's annual satisfaction survey from about an eight to an eight and a half out of 10. While we can't say for sure that the rights work was the sole reason for the improvement, we did get a few respondents specifically mentioning it. We've also had positive feedback from teachers at the recent Lianza conference. Excuse me. So what did we learn from the experience? For me, there were four main takeaways. Thing learned number one, always be positive. One of the best pieces of feedback that we received during the prototyping phase was from Mandy Henk of Toa Toa. She made the excellent observation that even for all rights reserve materials, there is always something that you are allowed to do with them. She pushed us to ensure that there was always a green light somewhere in our traffic light system even for the most rights encumbered materials. We initially trialed some statements that were a bit more prescriptive and specific about what one could do with copyright materials, even going as far as trying to explain a bit about fair dealing under New Zealand copyright law. However, we hit a number of exceptions across the different formats. For example, rules for audio were slightly different than those for film and those for images, and it was really hard to get a statement that worked for everything. In the end, we settled on this somewhat less specific but still positive encouragement. And by the way, that more information link initially led to a copyright explainer that was on a third party site. I have forgotten the site, I don't have it in my notes. But that site disappeared at some point while we were working on this shortly before our launch. We thought about writing our own explainer about allowed uses for copyright materials. But in the end though, we realized in our research that one of the best clearest and most definitive explanations about what you can do with copyright materials is the subpart of the actual copyright act that deals with what you can do with copyright materials. Go figure. So that's where it links today. That's legislation.gov.nz. That leads to thing number two. Words are good for explaining things. I don't know if anybody knows the book Don't Make Me Think by Steve Krug. It's still one of the best books about usability and user testing on the cheap. If you don't have a copy in your institution's library, go out and buy it tomorrow because it's great. This is one of Steve Krug's lessons is get rid of half the words on each page then get rid of half of what's left. I'm usually a firm advocate of the School of Thought that says fewer words on a webpage lead to more clarity. But there are exceptions to every rule and we occasionally encounter cases where more words can serve to add clarity. Our copyright block is definitely one of those cases. Through our user testing, we found that extra words reassured our visitors that they understood what we meant by share and modify and that the extra space in the design led us offer an example or two. This becomes an opportunity for passive learning. So not only does somebody see the big bright icons but they get an opportunity to say like oh that's what no commercial use means or oh if I put something on my blog post that's considered sharing. Good graphic design principles keep us from getting in trouble legibility wise the nice bright icons draw the eye to the overall value of each point and there will never be more than six icons showing which keeps us square with Hicks law. Hicks law if you're not familiar that's the research that shows how many things we can keep in our head cognitively at one time. So you might have heard you know don't have more than five choices in a menu or don't have more than seven choices. There's no hard fast rule but it's kind of a curve but that's Hicks law there's a great Wikipedia page on it. A bold headline for each part of the statement makes it easy to scan the text as well. We also made sure to add enough typographical space between each statement so that each reads distinctly. These same statements would have been far less successful had they just been written into a single paragraph. So speaking of legibility thing learn number three nice designs make the bad stuff more obvious too. Perhaps the most surprising outcome of this project was the sudden discovery of a number of problems lurking in our rights data. It turns out that if you improve the legibility and clarity of your data any previously hidden problems with that data also become clear and legible. Most commonly there were some cases where rights usage rights and rights URL held contradictory values. In some cases the errors were in the content partners original data and in other cases the problem was with our interpretation at harvest time. Either way the clarity of the new design necessitated a whole tranche of work to clean up these now highly visible inconsistencies. At the time we probably felt like this was a bit of a fail but now that we've closed a large number of these bugs it feels like a case where the new design led us to data improvements and a better service. And the last learning I'll share today is this reminder we don't need to be perfect we just need to be better. Our team runs an agile management process and preaches a philosophy of continuous improvement. This stream of work felt particularly conducive to that style of working considering that the design we started with set a frankly a pretty low bar for help helpfulness. We built the prototype yes and we tested it yes and we were able to learn yes but that in itself is not iterative design. The fact that we tinkered with this design on and off for about 18 months making changes big and small along the way is what really helped us have confidence that when we finally committed this somewhat tricky bit of logic to code knowing it still wasn't perfect we'd already tested our way through the everyday use cases as well as some of the more unusual ones. And speaking of iteration we're not done yet so what are what are some of our interesting problems still left to solve? A feature that we had in the early prototype but ran out of time to implement is a tool below the rights block that makes it easy to properly cite the item that you are looking at in a variety of academic and online formats we'd really like to add this soon. As an aside what's on screen is like the roughest of rough drafts and I'd really love to get folks feedback about what citation tools would be most useful as we head into 2020. The draft that I'm basing this on was probably written 10 years ago and I'm sure the latest online citation tools are totally different now. Our library also continues to work with the rightsstatements.org coalition and we're hoping that our input can bring their offerings more in line with the New Zealand copyright context so that we can get on board. Without boring you with the details one issue is the lack of an explicit public domain status in New Zealand as there is in the USA and Europe. Another issue is that there is not any internationally interoperable standard for communicating cultural restrictions. We've been learning from efforts like local contexts, traditional knowledge labels, for ways of properly respecting Tikanga Maori and other Pacific cultures, but ideally our ultimate sort of ideal case would be to expand the rightsstatements.org work to encompass this inherently rather than having to tack it on as an extra thing from the outside. We're also keen to take our learnings from the digital New Zealand version of our rightsstatements to our other properties, most notably the National Library website and papers passed. There's a lot of data wrangling work going on behind the scenes at the moment to support this. And finally we're starting to run some experiments with the new ways of measuring impacts. We've been looking at the Topapa audience impact model that Adrian Kingston presented at NDF last year and we're adapting it to fit our library. As part of that mahi we hope to be able to collect and share stories of how our rightsstatements have afforded deeper and easier use of the collections of our content partners. The digital New Zealand and National Library teams will be manning the booth in the Oceania room during all the breaks so if you would like to have any further deeper discussions on any of this stuff please do stop by and have a chat with us and I'll be happy to take any questions that you might have. Thank you.