07/21 - Flatout Friday!

I'd only be limited by the PRE cap, that pace is me taking my time, adding changes to the script, talking with people, taking breaks etc. I'm in no rush cause I'm stopping at 2k.

 

Yeah same here. PREs are what limit me here.

 

Well aren't ya'll special?

 

Eh, that pace isn't even that close to the PRE cap if you use both site.

 

Well I tried doing both but I had issues with www not submitting properly like it likes to do.

 

Anyways I finished as many of them as I'm willing to do with my buffer.

 

3/5 gone, will probably close soon even if short of 5 so there is enough batch left to work.

 

Does anyone actually know how big of a market research or whatever we do actually is? I'm surprised there's not more specialized software you can buy.

 

Title: Application to Complete a Checking Task ($2 when you apply, and another $12 if you are selected)(~ 20 minutes) | PANDA Worker: Preview | Accept | Requester Requester: Psychology [A35QM4BKNUJS0B] (Contact) TO 1: [Pay: 3.57] [Fast: 4.85] [Comm: 5.00] [Fair: 5.00] [Reviews: 34] [ToS: 0] TO 2: [Rate: $7.50/hr] [Pen: -- days] [Res: -- of 0] [Rec: -- of 0] [Rej: 0] [ToS: 0] [Brk: 0] Description: We are looking for some MTurk workers to complete a 20-30 minute checking task for US$12. To apply for the checking task, please complete this HIT first. The application process (i.e. this HIT) will take approximately 20-25 minutes and involves completing two questionnaires. We will remunerate you $2 for your time. Within four weeks, we will invite the strongest applicants to complete the $12 HIT (which will take 20-30 minutes). Time: 1 hour 30 minutes HITs Available: 1 Reward: $2.00 Qualifications: Inc: [40840-50881] Exists ; Exc: [51470-50881] DoesNotExist ; Location In US;

HIT exported from Mturk Suite v1.21.11​

 

Closing sales down at either 5/5 or 3.75k HITs. Looks like the batch has ~2.5hrs left in it at this rate.

 

I guess I don't understand how you can view and compare those a9s that fast.

Like, working flat out, I figure I can accurately judge one every 4 seconds, but between amazon load time and mturk submit/reload time I don't think I can do more than one every ~7 seconds.

 

Scripts yo

 

No, I mean the cumulative time it takes to:

read the product title ~1-2 seconds
look at the amazon page ~1-2 seconds
judge whether the former is similar enough to the latter to hit good enough or not good enough ~1-2 seconds

 

And I'm saying, scripts yo.

 

That's because we're https://www.reddit.com/r/totallynotrobots/

 

In fact, these are making me super nervous, I'm sure I'm going to get mass rejected and then I might as well quit doing turk at all because it will totally tank my approved ratio.

 

@LurkerLUL Priya Ganapathy just sent me an email about a 1 month followup thing. Check yo email. (Although it might be for one that you missed)

THANK YOU FELLOW HUMAN

That is what your worst case scenario thing is for. If you are scared (like I am lol) only do as much as you can do without going below 99%.

 

He made a script you can purchase if you haven't seen the advertising. His script probably opens the page inside the iframe, negating the need to open and view another page. So it's about 2-4 seconds per HIT, probably closer to 2

 

I mean, consider the time it takes to

A: read one string of text
B: read a second string of text
C: determine a relationship between the two strings


when you have been up for 22 hours so far