What is UP with the images/photos?
I am a really big proponent of visual learning, and I'm happy to see it making its way into more duolingo lessons.
But what is UP with these photos? At least on duolingo Italian, I feel like I've rarely encountered a photo that is helpful.
There was a photo for "the bear" in which the bear was almost invisible amid the surrounding image of a river. For "measure", there was a poor quality photograph of a door with a bunch of ilegible writing on it. "The law", a hallway. "winter" was a picture of a government building, albiet there was some snow in the image - but the other images were not anything to do with seasons or time.
The point of visual learning is to trigger an instant visual association with the sound of the word. But how can you do that when even in your own language, you'd have to make five or six guesses to know which thing in the photograph is the intended object?
Are these photos somehow randomly selected by a computer program, or does someone hand select them? Because it just feels like no one has looked at them. I would really love more emphasis on visual learning, but if it's going to be implemented in this way, or if there isn't currently enough budget to get good stock photos or contract a good photographer, I wonder if it might be better to leave it out for now.
Yes, I totally agree with you. Many of the photos do seem random, and barely connect to the word.
that's true. it's hard to decide what the word is when there's 2 objects in the photo
I often wonder how much of Duolingo is computer generated. My guess is that the developers are playing around with computer mechanisms to make sentences and pictures. If they do, it is probably not an option to hand-pick or even draw images for lessons.
Pictures I've seen in French and German courses were mostly good.
I think this could be fixed relatively easily by just adding an "up/down" rating system to the images, and allowing the community to filter them.
Yeah, there was one where it asked me to click "l'animale" (the animal). There was a picture of a dog with "il cane" (dog) under it, a picture of an elephant with "l'elefante" (the elephant) under it, and a picture of a cat with "l'animale" under it. ...Whaaaaaaat? I got a laugh out of it, though.
My favorite is the image that was supposed to be of a mouse (as in the rodent) that had a dark background, an oval shape, with a red body. It also had a bar-code on the body.
It was, in other words, the underside of a Microsoft Mouse.
That's when I realized the images truly must be computer-chosen with no human vetting!
I think DL should make some custom drawings. I would prefer like a cartoon of it, but then again you can get too accustomed to the cartoon image and forget the real item... So a nice clear photo would be nice, and if it's something vague such as ''winter'' then it should be made extra clear e.g. cartoon snowflakes, snow, and more, and it shouldn't be able to be confused with similar things like snow.
I, however, haven't had a problem with the images. Well, I found a few to be a bit unclear but it wasn't too bad. I don't learn Italian much so I don't know, but I can see why this is causing a problem.
Same with me. I am remembering my words fine with the pictures. Even the unrelated ones. If i thought of law then i would think of the hall and be like, oh that hall is a courthouse.
Some of the photos are okay, but they haven't helped me remember many words. Maybe the process of picking out a photo out of 4 or more would help out.
Would be awesome if we had some sort of uploading and voting system at each word webpage. LIke Memrise.
I don't think you are right.... they generally understand that the image may not be very clear and provide you with a hint.
The thing is, though, that if you've ever tried at least a unit or two of Rosetta Stone (who's primary mode of learning is visual) you probably know how incredibly effective it can be if done correctly. The point of this method of learning is not really to provide a hint to the meaning of a word as much as it is to create and trigger visual memories.
I generally find that I enjoy Duolingo more than Rosetta Stone as a learning experience, but I think the visual component needs a bit more attention.
Yeah, I've also done two of three Pimsleur phases (which amounts to about 30-40 hours) and several "units" (or whatever the complete lesson packages are called) of Rosetta Stone.
I completely agree with you, I prefer Duolingo, and that's kind of why I'm pining to have the visual approach improved. As a learning experience Duolingo is far more enjoyable (and thereby probably more effective) than both Pimsleur or Rosetta Stone, but when it comes to the effectiveness of the image-based component - objectively - Rosetta Stone is far superior.
Rosetta Stone, to be fair, has a different approach to learning. They never give you an english translation, for one. You're supposed to learn the meaning through visual association and context, and speaking is a big component. I'm not suggesting Duolingo change their learning mechanism. But ultimately, when it comes to the difference in the effectiveness of the images - what they do in RS not that difficult. They just have really good quality images that convey their meaning simply and fairly immediately either directly or by the context of images surrounding them.
I'd love to have more visual learning, but only if it's done right. I think part of why I wouldn't like Rosetta Stone as much is because there's absolutely zero translation. So, sometimes you don't even know what a word actually means. Could get difficult when you're learning about words for abstract ideas and such.
Exactly, I think "well rounded" is the key.
Each of the tools I've used is lacking in at least one key area. Duolingo is, so far, the most well rounded, and keeps you coming back. I never fully understood the grammar component, for instance, until I started using duolingo. Though I will say, regarding your concern with RS, the zero translation thing can be challenging at first, but after a while it becomes extremely effective at helping you truly grasp the "meaning" instead of sort of mentally "translating" - at least for some learners. But that may depend to some degree on whether your learning style is visual, auditory, reflective, impulsive or some other type.
At any rate, I'm certainly not advocating that duolingo change their approach to learning - just give a little more attention to what they're doing with images, providing contextual cues and more clear, descriptive images.
Yeah, I agree that learning without translation can be effective, but I think RS probably is best used alongside something else (e.g. a Spanish-English dictionary) to make sure you don't learn a meaning incorrectly. I haven't technically tried RS, but I've watched someone else use it, and I don't think it's right for me. Duolingo is great, and even if Duolingo was majorly lacking in some area, we'd still be getting more than we pay for!
I've tried Duo's French and German. I find the German vastly superior because it has actual grammar lessons. I discovered them by accident: I was stuck on some unit 5-6 units in, was very frustrated, looked around the page, found the grammar, and then went back and read it and the preceding ones. The amount of grammar explained is just right, and it helps to be able to read it after having struggled a little.
The French duo has very little grammar at all, and I find that much more frustrating. So even within Duo, there is variation.
I agree that an intelligent use of images, along with the German Duo's levels of grammar, would be a huge win.