Title: Survey for baseball fans – 10 to 15 minutes for $.50 plus plenty of chances for bonus pay. Requester: The Wharton School [ABIUW8FBC3EE7] (TO) Description: 5 to 10 minute survey for $.50 Reward: $0.50 Qualifications: HIT approval rate (%) is not less than 90, Location is US Link: https://www.mturk.com/mturk/preview?groupId=2KGW3K4F0OHO2MT806BOAL97INF10G love this one.
Yeah I do that too but it seems some review are only on Name + zip and then if nothing sometimes it shows up in just address. It's just annoying and it messes up any flow I might have.
ok, i'm not sure if it's b/c i'm drunk or b/c there's nothign to work on, BUT OMFG Y'ALL ARE FUNNY OTNIGHT! ok, i swear i'm going ot bed now. Finally talkt ot the bf, he's doign good. f*ck fixing typeing errors, "m drunl y'all!!! WHEEEEEEEEEEEEEEEEEEEEEEEEEEEEE
It's just too odd. Either Canada has not put much effort in putting their restaurants online or they way they do things. IDK! It's just freaking annoying and it makes me a grumpy bear. Then I go slower... and well its all downhill. If you can't tell the hits havent moved much. IT was up to 14k. Started around 4 Pm I think. That is really slow movement.
There and I think I managed to not mess it up http://mturkforum.com/showthread.ph...c-HIT-s-9-13-Foxy-Friday!&p=454747#post454747
Hahaha. Yeah I don't have a stone but that same principle is what got me burnt. I blasted my oven as hot as it would go with my pan in it. I had to put an oven mitt inside another oven mitt in the end.
I try to be efficient, because I know it's important to a LOT of folks on mTurk... but it's also a resource issue - there's just one of me, and I have to be sitting at my computer to run the grading program, then I have to go through and review any that are flagged (this can be fast if almost everyone does a great job, but sometimes it's a lot of slogging through ones that obviously are random or unconsidered, and having to reject 'em). Once that's done, I process the bonuses - usually within a day or so, but sometimes immediately if I happen to be available to do it at the time
I ran this through the automated sentiment analyzer I'm programming, and it came up as "[4] Extremely Positive". I can feel the love, and I appreciate the props...
It's how close you match the 'wisdom of teh crowd'. For example, if 20 people classify a text snippet, and the distribution is something like: 3 people classified it as “Extremely Positive” 9 people classified it as “Very Positive” 5 people classified it as “Moderately Positive” 2 people classified it as “Slightly Positive” and 1 person classified it as “Very Negative” The grading program flags the answer. It's okay to get a flag here and there (because it is subjective, after all)... Also, you have to get at least 8 out of 10 of the golden items (where we already know the acceptable range of responses ahead of time). Next, if you're getting more than 4 flags on the remaining 20 items where we don't already know the acceptable answers, you probably won't get the bonus. Finally, if you're getting more than 12 flags on all 30 items for a HIT, I have to go and manually review to see if the responses are just random/unconsidered to check whether to reject it.
Question please. Why is it that sometimes when I log into Amazon Turk, I check my account and then go to hits. It seems the screen flips or something. It looks different and won't let me access the hits. Its a blue color with funny edges. Am I loosing my mind or is my computer being taking over by aliens?