They seem pretty safe. Same requester posted back in July and lots of people did them - I didn't hear of any rejects. Do a good job and you probably don't have anything to worry about.
Excellent. I'm at 600 and really trying to break 1,000 by the end of Monday. This should help Thanks!
Switching between Productrnr and Gandolph is really working out nicely. Productrnr approves within 3 days, small batches pops up here and there.
If you carefully read/parse their answer, you'll see they suggest three different areas in which you might have run afoul. Note to everyone else. This is just a generic list of possible issues, not a specific commentary on the actual HIT. Or any other actual HIT. 1. "The rate at which you provide answers"--This is most likely a measure of simply how long you spent on hit from "Accept" to "Submit". If you completed the HIT first, then accepted it, and then immediately submitted the code, that would set off alarms. Or if you completed the HIT in a less time that the requester thinks is reasonably possible, they might reject. A good number of requesters automatically reject everyone that finishes before at set time, and a handful will even nail anyone who finishes in the fastest XX% percent. 2. "how well you follow instructions"--Most surveys have at least one "Attention Check", aka "AC". The simplest type is a long complicated question with the instruction, "Ignore this question and do not answer it" embedded somewhere in the verbiage. Or there could be a long list of items, and the hidden instruction will tell you which one/s to click. If you miss this sort of AC, you're assumed to be going through the survey too fast and/or failing to properly read the instructions. 3. "how accurately you respond to multiple questions of similar kind"--This is another form of AC in which they repeat the same question several times, to see if you give the same answer. If you don't, you might be just filling in questions at random. I'm not necessarily suggesting that you are at fault here. Many, many people on this forum run afoul of #1 on a regular basis, sometimes justified, sometime not. Type #3 ACs can be quite subjective, and type #2 ACs can be cleverly hidden. Finally, there are always HITs with programming errors, and never rule out the possibility of requesters simply deciding to scam you. It happens. Make sure you send an email to the requester to discuss the situation. If you are polite but firm, they may be convinced to reverse the rejection. Also, please leave a review of the requester on Turkopticon. Thank you.
https://www.mturk.com/mturk/searchb...rce&minReward=0.00&qualifiedFor=on&x=20&y=13# Averaged about 20s each. Can repeat 16 times before being soft blocked. After a few you'll see why these are super fast.
You're not supposed to highlight them. You're supposed to type them in. People here have worked out tricks to allow highlighting/copying, but I don't use them. The HITs are designed to trick Google into listing certain search queries as more popular than others, and when you actually type the search terms into the box, Google gives them a higher weight than cut/paste entries. Supposedly. If you read more about the people mentioned in the HITs, they invariably have (or have had in the recent past) serious problems going on that they don't want to be immediately obvious. Every time you type in "John Doe Lawyer" it becomes less likely that "Joe Doe Convicted Pedophile" will show up in the search results. Hence in part the need to type the term in by hand. It's a long story, but you have the basics. I'm actually working them right now. They are among my favorites, even though they don't pay very much. Oh, and they take me a minute and a half each, assuming I manage to prevent myself from reading about their latest predicaments. Some of these guys I've been whitewashing for nine months now.
My bad guys, I meant the $.16 ones named "Search: 8 Keywords on Google.com (US)" The others are relatively decent paying.
Unfortunately I can't discuss HIT contents, but after 2-3 you'll figure out the trick to doing em really really fast.
Most aren't that great, decent to do when there isn't much else to do, but the ones I named above are an exception.
Mediocre TO, small amount of writing, took me less than 4 minutes - I did one for him before and it paid quickly so I wasn't worried about the TO Title: 5 min. purchase experience surveys Requester: Kevin [A2XS7WLMVUD3RF] (TO) Description: Give us your opinion Reward: $0.30 Qualifications: HIT approval rate (%) is greater than 95 Link: https://www.mturk.com/mturk/preview?groupId=2ZSQUQIHPCGM5KWFMMN56H0LQWX06I [size=-2]Powered by non-amazonian script monkeys [/size]
It's not a soft block. It's an internal limit each person has on crowdsource's end. They dont "softblock" through amazon. They limit each person because there's only so much one person can do to effect SEO through that method.
I don't think you need to worry about discussing the ins and outs of this particular type of HIT. Going on about the contents of the survey HITs is what makes everyone unhappy. Discussing batch techniques is fine. Crowdsource hardly ever rejects, and they virtually never reject for this sort of HIT. But I have known one person who did manage that feat, so beware, they are not inattentive.
HIT contents are fine to discuss. It's survey content that can't be discussed due to them actually stating to not discuss them when you consent.
OHHHHHH. Well then in that case, the HIT is the same task repeated asking for a link for the same 8 search terms. I just opened notepad to copy down the links, and copy pasted them into the HIT.
...You did perform all 8 of the different searches, right? Because if you just paste in (for example) http://www.rosicki.com/ eight times in a row, that might upset them. That would let you finish in 20 seconds, but, as I noted, once in a blue moon, they have been known to reject for precisely that.