@westboundwarbler – Interesting and useful research. Thanks for doing this.
5 Likes
The real number of users is closer to half a million. That’s using the fairly generous metric of having posted or IDed anything in the last two months. If you use last 30 days, it’s more like 300K.
That said, a sample size of 26 isn’t going to tell you much. The real problem is we know nothing about the 26 people. Are they active users? Casual users? Have dozens of observations? What countries are they from? You’d need about ten times as many to even begin to cover major variables. And that’s not even accounting for the fact that the choice to even respond to the survey introduces a bias. We know nothing about what the other 78% percent think.
3 Likes
Well, the target population was not all users. I only sent surveys to users who had all rights reserved. And I am not sure what proportion of users have all rights reserved. Finding out this proportion is the subject of further research for me.
Does anyone know what proportion have all rights reserved? or any CC license for that matter?
2 Likes
All observers surveyed (n=114) had been active within one month, to ensure a higher chance they would answer my survey.
All observations made from North America (U.S./Canada). Number of observations per observer ranging from 55 to over 64,000.
Good thoughts and I am happy to answer any more specific questions.
4 Likes
some of the statements you’re making here seem a bit flimsy based on “preliminary findings from a qualitative study”. a lot of what i’m seeing sort of sounds like you went into your project with a particular angle / goal and are just trying to find whatever information you can to justify it.
if you’re going to share research findings, i think you should post your paper / draft and let folks comment on that, preferably after you’ve reviewed these with an advisor or some research peers. putting this into a forum post instead seems like a backdoor way to fish for research help or to promote an agenda.
3 Likes
Since you restricted yourself to US/Canada you can only make conclusions about US/Canada users. You can’t say anything about the license preferences of users in Europe, Asia, or anywhere else.
You also need to address any discrepancy between the surveyed population (n=114) and the population that responded (n=26).
The survey size is still far too small.
I think the study has more fundamental problems.
If I’m reading this correctly, you’re saying that you did a major study on license choice without even getting basic data on the frequency of license choices? That should have been done before sending out any surveys. In fact, it should have been done before writing a survey. If you don’t know how many people use the restrictive license your initial claim that there “is a barrier limiting the full potential of iNaturalist contributions” involving the choice of license can’t be verified. The whole study could be for nothing.
2 Likes
This is a very good point and a great suggestion. And I want to add that I think any time licenses are mentioned at all it should be made clear there are separate choices possible for observations and sound recordings and photos – I don’t think this is very clear as it is presented now.
1 Like
This is a good point and I think it is part of the problem. I think the checked box reduces the license choices to something that looks like a ‘yes/no’ decision when it isn’t. There are nine choices and can choose differently for observation, sounds, or photos. The way it is presented makes the choice look like a ‘yes/no’ decision without telling the users what ‘no’ or unchecking actually means.
What if the default to an observation was CC0 and CC-BY-NC for photos and sounds? That way actual intellectual property has attribution but the observation is very open. It has been widely debated whether or it is even reasonable to reserve rights to an observation. https://forum.inaturalist.org/t/does-copyright-law-allow-you-to-restrict-use-of-your-observations-the-data-this-is-not-the-photos/41709
2 Likes
I am confident in my research. My advisors, peers and I all think the findings have been very interesting, relevant, and worth sharing. And, the users who have all rights reserved by mistake also think this is worth sharing.
When users learn that they have mistakenly reserved all rights they are very appreciative of the information. Many expressed gratitude to me for doing this research. These users do not like the fact that they made this licensing error and have shown to be eager for the information they need to choose a CC license once it is brought to their attention. In this survey, I did not ever suggest that anyone change their license because I posed this study objectively. However, when these users learn of the mistake, they do go into their account settings on their own to learn more and make a change. Between the 15 respondents who mistakenly reserved all rights to their observations, if all of them adopted a CC license that shares with the GBIF, that would share a total of over 28,000 observations (observation they intended to share but were not). A more systematic attempt to bridge this gap would be well worth it. The take home message for here is that if we bridge this information gap, it will make a difference, and the community will appreciate it.
4 Likes
if you just wanted to make some suggestions about how to improve the system, no one would object, but if you’re going to say that your suggestions are backed by research, you really should provide a way for others to review that research. if you’re confident in your research, i really don’t understand why you wouldn’t share the actual paper.
4 Likes
@westboundwarbler – I think you have done a good preliminary study and made some of its limitations clear. As always, there is much more we’d like to know and some of those things strongly influence the results you get. That doesn’t mean your work isn’t good, just that it doesn’t, can’t, go as far as we’d like.
One thing we need to keep in mind is the with surveys like this, the number of people who respond is always frustratingly low. And often that subsample is biased in some ways. (Talk to political pollsters about this!) Oh well.
What you have done that’s important is find out that a lot of the people who have restrictive licensing in place didn’t intend that and don’t actually want it. (That’s true no matter how biased or unrepresentative your sample may be.) That brings to our attention a problem that perhaps can be solved, or at least reduced.
4 Likes
Your comment is true…but for people like me who have used iNat since 2017 this all seems confusing. I have spent the last half hour trying to understand what I want to do…I did find the section that explained the 8 differences, but didn’t ever see one that says “License my photos, sounds, and observations so scientists can use my data.”
3 Likes
Hello fellow Nebraskan! I enjoy seeing all of your pollinator photos!
When you click the drop downs to select a licensing option, they have little “GBIF” and “Wikipedia” icons next to them. The “GBIF” icon essentially translates to “let scientists and researchers use this”. It let’s others use it too though, so be mindful of that.
If I remember right, you embed a copyright logo into your photos. You might consider changing your Observation license default to “Attribution + Non-commercial” but leaving your photo and sound license defaults to “All rights reserved”. This would allow scientists to see that pollinator X was seen in Nebraska on such and such date, and any notes you put on the observation, but they wouldn’t get the photos. They’d have to navigate to the observation in iNat to view the photos, where you’ve reserved (for yourself) all rights to said photos.
To your point though, yeah, it takes more time and energy and understanding than it should in an ideal world to figure out which is the right set of options for ones needs.
5 Likes
that’s becasue there is no single license or set of licenses that magically makes your data available for scientists to use.
while it is true that there is a set of observation licenses that allow research grade observations to flow to GBIF, and there are certain photo licenses that allow photos to be included in the AWS open data set, and Wikiworld likes another set of photo licenses, etc., the reality is that scientists already have access to pretty much all the stuff in iNaturalist itself. they can already use that stuff in a lot of ways without worrying about copyright / licensing. it’s only when they want to copy / distribute that stuff or derivatives of that stuff that they have to be mindful of the copyrights and licenses.
other than choosing a CC0 license, which effecitvely allows anyone to use your stuff for any purpose without seeking your explicit approval, no other single license will allow every scientist to use your stuff in every way that they want because every use case is different.
so if you don’t care about anyone using your stuff for in any way, for any purpose, without your explicit approval, then go CC0 for everything. if you want to explicitly approve every use case, then you need to reserve all rights (no license). in between that, iNat offers a good default license that allows for many (but not all) scientific uses without your explicit approval – non-commercial, with attribution (BY-NC) – but you can change that to whatever you prefer.
copyright and licensing are not easy concepts for many people to understand. there’s not some quick 5-minute way to train someone who knows nothing about copyright so that they will understand all the nuances of it. so if you really want to contemplate something other than the 3 choices i mentioned above (CC0, CC BY-NC, no license), just make sure you spend some time to read about and learn what those choices mean. there’s no other way to make an informed choice.
10 Likes
Thank you for your prompt and easy to understand directions. I also just figured out how to add the different option. You just click on the tiny little arrow…I have attached an image to show how I did that.
I do put a copyright on my images…even though I am not worrying a lot about people using my images and making millions of dollars! :)
Thanks again for your clear and concise answer.
1 Like
Thanks for the prompt and informative answer. I guess I always thought the purpose of iNat was to share information among the scienctific community. I think your reply plus the one from regnierda really helped me understand. I have a basic understanding between copyright and licensing…but now that you helped me understand that scientists have had the ability to access my data all along, eases my mind.
2 Likes
I do not have the info for observation data licenses, but the proportion of users by image licence used is:
CC0 0.15%
CC-BY 0.28%
CC-BY-SA 0.06%
CC-BY-NC 52.32%
CC-BY-ND 0.01%
CC-BY-NC-SA 0.09%
CC-BY-NC-ND 0.09%
All rights reserved 47.00%
Users are roughly split between “all rights reserved” and CC-BY-NC, with less than one percent of users having a different image licence than those two.
Further details can be found at a journal post here
This snapshot was taken in Jan 2024 and has remained relatively stable over the past few years, even with the minor increase in the use of CC- licenses following an active campaign to have users change their licence to a CC-license in order to reduce data storage cost for iNaturalist.
1 Like
The breakdown of observers that have submitted audio recordings has a much different similar proportion:
CC0 0.37% 0.9%
CC-BY 1.08% 1.8%
CC-BY-SA 12.73% 0.2%
CC-BY-NC 0.16% 52.9%
CC-BY-ND 0.02% <0.1%
CC-BY-NC-SA 0.25% 0.4%
CC-BY-NC-ND 0.24% 0.3%
All rights reserved 85.15% 43.5%
nb. edited to reflect all sound observations.
Wow, that’s a lot of all rights reserved… I’d have guessed it to be closer to like 10% or less. I wonder if all of those users explicitly chose that licensing option, or if and some other factor accounts for such a high percentage.
your numbers here look wrong to me:
- there are 62.9K observers who have recorded sound that is licensed based on: https://www.inaturalist.org/observations?place_id=any&sound_license=cc0,cc-by,cc-by-sa,cc-by-nc,cc-by-nd,cc-by-nc-sa,cc-by-nc-nd&sounds&subview=map&verifiable=any&view=observers
- there are 111.2K observers who have recorded sound based https://www.inaturalist.org/observations?place_id=any&sounds&subview=map&verifiable=any&view=observers
- 62.9 / 111.2 = 0.566. even if you say that some observers could have both licensed and unlicensed sound, that shouldn’t move the needle that much.
so unless you’re somehow able to see actual users’ default licensing choices rather than just inferring them from observations, i think you might need to take a second look at your numbers.
as far as i know, this can only be an approximation, since i don’t think regular users can see what a user’s default license choices are and can only infer them from actual observations. approximation is fine, but your journal post seems to characterize your numbers as actual default licenses, which isn’t quite right.