Why isn’t there a NatureNerd Trivia Quiz Game? Or is there?
iNat data would make for a great natural history quiz app. The trick would be to empower users to create their own quizzes. User creates a target list of taxa, for example 16 common trees. That could be the extent of the customization: just create a taxon list. Fancier options could be to restrict images geographically, choose from some game play options, or whatever. Each turn, the game fetches a random image from Research Grade Observations of one randomly selected taxon from the list. Player must select correct taxon, or select from 4 choices that include the correct (matching) taxon + 3 incorrect distractors. Game continues, tracking percent correct choices. Could be chunked in sets of ten images.
Teachers, students, and nature nerds would probably enjoy this. The ideal thing is that it can be very easy (10 common and very distinct bird species) or very hard (20 devilishly similar species of plants or fungi) and therefore can be useful to many different audiences.
I think this app could be widely used. It would be so much more fun than the canned quizzes that have only a few images of each taxon and can only be played once or twice. I do not have skills or time to work through the development process but if I did, I’d do it. There were the beginnings of something like this using Encyclopedia of Life data at one point, but it was lost in the general implosion of that entire website. Thoughts?
I’m imagining this something like the eBird quiz - maybe this could even be a feature within iNat? I know, it would probably be hard to develop (I know nothing about programming, so please correct me if I’m wrong), but I would find this a pretty cool feature. It could also be used to learn how to ID certain species before going to a new place, etc.
Yes, a lot like that - except customizable to limit the taxon set. Even in a given location and season (as done in the eBird Quiz) there would be a frustratingly huge range of possibilities.
It would be cool as a taxa based thing, where it could show “This is Family F, and it contains genus X, Y and Z… X has… Y has… and Z can be distinguished by… here are some pics showing the features… and here are a bunch of images for you to have a go at putting to X, Y or Z”
We have most of this in iNat, with the identify pages and the about taxa pages… but a more “classroom activity” oriented version of it would be cool. I would imagine the makers of classroom resources could do a lot with this sort of interconnectivity with iNat.
unless you want something super fancy, i think the hardest thing is not the coding. it’s figuring out a proper workflow / design that makes sense. to do a quiz the way eBird does or as you guys have proposed, i think the first thing to consider is the quality of the photos. how do you ensure that all the photos will be relatively high-quality for the quiz? do you make a project(s) that contains observations and source the photos from that (extra work)? do you pick from the curated photos on the taxon page (limited selection)? or do you assume that photos could be relatively low quality? if photos could be relatively low quality, i think that probably means that you should consider approaching the quiz in a different way – maybe more like one of those 3x3 or 4x4 CAPTCHA grids that asks you to select all the buses or traffic signs that you see, or a tinder-like format that asks you to swipe right on only, say, cardinal photos.
also, do you incorporate geography into the sourcing of the photos and possible species?
I envisioned limiting photos to Research Grade. Poor photos (and yes, some misidentifications) add to the challenge but don’t make it less useful so long as taxa are chosen with care (e.g. relatively easy-to-ID based on photo evidence). The whole idea is NOT to use just one or a few known photos but instead to fetch from Observations. It’s harder to code but the correct tool would be a quiz that makes itself new each and every time it is played. The other option requires more time to make the quiz upfront and it becomes useless after one or two plays.
“Easy to code” is very hard for a totally ignorant person who is not interested in learning those skills. So if you think it would be easy, by all means put something together! It may not be exactly what anybody else envisioned but it would be a start. Or if you really want design specifications I bet the people here could knock out a list.
For me it would be to go back to the starting idea. People can both Build and Play quizzes. A “Quiz Builder” chooses a limited number of taxa (list of possible species). The Quiz software then automatically fetches a randomly selected image from an RG Observation of any of the taxa on the list. Quiz presents Player with the image and below it four taxon names: the correct answer plus three distractors from the list. Player wins a point by choosing the correct name. 10 presentations could be one quiz, with extra points for fast responses.
I think a geographical filter would be useful but not strictly necessary. The Quiz would work and be easy for beginners if the builder chooses a list of say 20 common and distinctive species. It would be fun for expert users if the builder chooses 30 similar sedges or something awful like that. In my view the game itself should be very simple; the flexibility is in the list of taxa chosen by the builder.
i still don’t think it makes it harder to code, but it makes it harder to play, i think. below are some examples of some woodpecker observations. don’t read the comments, and see if you think it would be easy to pull the right species from a list of 4 choices:
i’m just thinking if i were a player, i would only have to come across a few of these kinds of photos for me to decide that the quiz game is broken or too tricky.
that’s why i suggested the CAPTCHA- or Tinder- style play if it was possible that there were going to be bad photos. but then again, that kind of a game isn’t that different from just pulling up (Needs ID) observations for a specific species from iNaturalist and agreeing or disagreeing.
things get even trickier if you need to account for geography or multiple photos, or other factors…
This app sounds like a fun thing to do. I have the skills and capacities to start a test version of this.
However, a feature such as being able for a user to make a quiz for other users would require databases and servers which would likely cost money to have.
So, in essence making this project would cost money, somehow these costs would have to be earned back. iNaturalist does not allow access to data for commercial purposes (although with exceptions maybe possible by permission). Also, I do not see a lot of traffic in this forum post which makes me wonder if there is a general interest for this.
Probably only a tiny fraction of the iNat community ever looks at this forum, so I wouldn’t deduct disinterest from few replies here.
About money: In the development phase a very small server should do, not that expensive these days if I’m not mistaken. I would be fairly optimistic of finding a monetary solution later if such a quiz goes live and becomes popular.
Wondering if the potential educational value of this proposed fun thing/ tool would fit in the parameters for any environmental/ biology/education grants out there (thinking of the many people involved with iNat that may have fingers on funding pulse). No sense whatsoever of current grants available, technical anything involved in creation and distribution, financial necessity.
the way i was thinking i would approach a web version of a game was to pass a list of taxon_ids via a web url. then that could be bookmarked, or you could use a url shortening service to remember that url. you could probably pass a similar list of parameters to a mobile app via deep linking. it’s not super elegant, but i think it would eliminate the need for databases to store user profiles / configurations.
For an example of a fun and applicable quiz game regarding plant ID, see https://www.birdandmoon.com/poisonivy/
Obviously only applies to certain plants in one region of the world, and may not distinguish radicans and rydbergii (but neither does your skin).
Embarrassingly, I missed 2 on that quiz while trying to do it as fast as possible :)
Thanks @pisum for your thoughtful response. I agree there would be some bad photos, but if limited to RG (and depending on the taxonomic group) it wouldn’t be terrible. There could be a simple disclaimer: Images are chose at random. Some will be tough to ID and in other cases the ID might be wrong!
Luck would come into it. But the game would not be terribly broken I don’t think. It would still be engaging and a way to practice skills.
@jtklein Thanks for taking the time to think about this! I think it would be ideal to maintain a database so that users could track their progress and all that fun stuff. But I am not ready with funds.
I work with students on identification of common species as part of a conservation ecology class I teach at a community college. This game would be a way to help them learn. I also know from experience that quiz games can be really addictive, and that the key is to have a very large set of items so that you aren’t repeating the same ones too often. A “new every time” iNat quiz would definitely have that quality and I suspect people would play if the thing existed.
@pisum I have no idea whether or how this would work, but if there is a way to do it without servers that is one option. Another option might be for me to approach the IT people at my college and ask them to host a modestly sized database for the game. If there is some light coding or implementation that could be done by our IT students, that would be a fun project that would further goals in other departments. So we might be able to do this in-house without a lot of external funding, especially if helpful folks created the basic structure for a beta version of the game!
Man, worst case scenarios for sure on those RHW images! I just now checked Recent Observations of Red-Headed Woodpecker; in a sample of 23 images, 3 were not immediately identifiable. Not sure if they are (or will be) Research Grade. This is going to be far worse for birds and mammals and other harder-to-image groups. So based on a terribly small sample I’d guess the “bad photo” rate is under 10%. Then at worst you’d expect 1 or 2 out of every 10 images to be something frustrating where the best you can do is guess. That wouldn’t be a terrible experience unless you are a perfectionist or a thin skinned person who wants a high score as validation of your expertise.
Important caveat for any teachers however: game is a practice tool only and should never be used as a summative assessment or “test.” For that you want to use hand-picked images and can use any of several quizzing or test apps such as Quizizz or your favorite LMS or a piece of paper and a projector.
@jtklein What else do I need? Expertise. I’ll reach out to the faculty who teach our coding / IT / web design classes and see if anybody would like to adopt this as a class project or independent study.