The second round of sample HITs went well like the first time. I'm looking through the answers and I'm seeing a some blank answers for the radio button group despite it saying in the instructions to not leave it blank. I've tried to figure out how to make the radio button group require at least one option picked but for some reason it doesn't go through with that property selected.

To be honest with you, I'm starting to not see the point of including it in the first place but my research team ultimately decided to include it. I'll make further discussion with my team to see if we should include it in there especially if I can't make a requirement to put in an answer for it. Just looking at the results, it looks like I might have to reject a bunch of entries because I was being too nice before. But it seems I'm getting some workers who's just skimming through the tasks and not reading the instructions. I'll accept those who look like they've tried to answer honestly. Next time, I'll see what I can do about that and raise the requirements.

Thank you for those who've done my tasks, especially those who have answered honestly and not just ran through it for quick cash. You will be rewarded. I have read some comments. Many of them were positives. I have heard complaints about unclear pictures. I know. I'm afraid I can't do anything about the pictures themselves. My original idea was to use one picture at a time per set, but my team went with this collage idea so I did what they asked. I'm not trying to throw them under a bus, but I'm just doing what I'm told. I'll see what we can do about pulling individual images rather than collages.

Again, I have no idea how people are finding my stuff but I can't find my own HITs lol

Edit: It's been brought to my attention that the reason for some blank answers is that sometimes MTurk deselect people's answers randomly. If that's the case, then that makes sense. After hearing this, I'll approve of everybody's answers again. I've been suggested to make the two questions two separate HITs. I'll probably do that. Thank you guys for your help and for doing my tasks.

 

Right now, I'm accepting and rejecting submissions but for the next batch I plan on sending, I might check for accuracy of themselves but I need to see if I can handpick certain workers because there are really some workers here that have been consistent with their answer quality. If you are one of those who answered my tasks honestly or you know how I can handpick certain workers to do tasks and if you're interested, we can make it work.

 

Why don't you set up a qual hit or test to designate a worker group, then place a limit on completion for one batch to see if any need to be weeded out? That way you'll be able to have a pool of trusted workers rather than having to weed through things every time.

 

I haven't been able to find them either, so don't feel to bad lol

 

.

 

.

 

I've been trying to figure out how to do that and get that set up.

 

Oh, that makes sense. Well, after hearing that, I'll approve everybody's answers still then. It seems like this radio button group thing is making it a hassle for both me and the workers. I'll talk to my team and ask them if I can get rid of it or something.

 

Could you possibly split the project up into two batches, then each batch would only have one question and you wouldn't have to deal with the radio button problem?

 

That's what I was just thinking. I'm talking with my team now to let them know. I'll probably do that. Seems to be the most logical solution because I've tried everything. I'm probably doing things wrong. I'll admit, I'm not a good HTML coder but for some reason I've noticed some technical problems that's likely not my fault with MTurk's toolset. There's an option with the radio button group template that it literally says it requires at least one entry in the group or it can't be submitted. I put that on there but it doesn't uphold those conditions.