03/03 - Soft Sunday!

It saddens me that you don’t appreciate my comedic efforts.

 

Sihao Chen [A2DMTTK9FYFGS5] Grade School Science Exam Question Answering - $3.75 | PANDA

Generous

Unrated

Unrated

$37.19 / hour
00:06:03 / completion time

​

Pros:
Super easy.

Cons:
Some writing depending on answers.

Qualifications: Location In AU, CA, GB, US; Total approved HITs > 500; HIT approval rate (%) > 95;

To read Sihao Chen's full profile check out TurkerView!​

 

Sihao Chen [A2DMTTK9FYFGS5] Grade School Science Exam Question Answering - $3.75 | PANDA

Generous

Unrated

Unrated

$41.03 / hour
00:05:29 / completion time

​

Pros:

Cons:

Qualifications: Location In AU, CA, GB, US; Total approved HITs > 500; HIT approval rate (%) > 95;

To read Sihao Chen's full profile check out TurkerView!​

 

Title: Pick the most sensible how-to guide or caption ending [ $ | $ ] PANDA
Requester: Rowan Zellers [A2RVUSQQ0150J2]
TurkerView: [ $12.13 / hour ]
Description: Given a few sentences from a how-to guide/caption, choose the most sensible follow-up sentence(s).
Duration: 01:00:00
Available: 1,036
Reward: $0.42
Qualifications:

• Total approved HITs GreaterThan 1000,
• HIT approval rate (%) GreaterThanOrEqualTo 98,
• ClassifyAIStuff1 Exists [ Request | How to Use ]

[375MF9UHMRXMAIPY38CNE17QJ03SXV]​

 

Sihao Chen [A2DMTTK9FYFGS5] Grade School Science Exam Question Answering - $3.75 | PANDA

Generous

Unrated

Unrated

$53.36 / hour
00:04:13 / completion time

​

Pros:

Cons:

Qualifications: Location In AU, CA, GB, US; Total approved HITs > 500; HIT approval rate (%) > 95;

To read Sihao Chen's full profile check out TurkerView!​

 

Sihao Chen [A2DMTTK9FYFGS5] Grade School Science Exam Question Answering - $3.75 | PANDA

Generous

Unrated

Unrated

$24.82 / hour
00:09:04 / completion time

​

Pros:
good pay. YOLO. I did google some of the questions trying to get answers correct, but...they say they won't reject as long as it doesn't appear the answers are just random, and I know mine were consistent...but they could have been consistently wrong.

Cons:

Qualifications: Location In AU, CA, GB, US; Total approved HITs > 500; HIT approval rate (%) > 95;

To read Sihao Chen's full profile check out TurkerView!​

 

Derek J Koehler [AN8LXOVUSVKGB] The Donor is in the Details(~ 7 minutes) - $0.75 | PANDA

Generous

Unrated

Unrated

$41.54 / hour
00:01:05 / completion time

​

Pros:

Cons:

Qualifications: HIT approval rate (%) >= 95; Location In US;

To read Derek J Koehler's full profile check out TurkerView!​

 

Did Rowan post a new qual that I missed?

 

He’s good! Cranky and aggressive to strangers as always, but he spends most of his days eating and sleeping in sunbeams and basically velcroes himself to Matt when he’s home- he’s the happiest little demon kitty now

 

Title: Identifying relationships between concepts [ $ | $ ] PANDA
Requester: Hannah Rashkin [A5ZY2NLV5IM1Z]
TurkerView: [ $11.61 / hour ]
Description: Identify which words and phrases can be used to fill in a particular relationship
Duration: 00:30:00
Available: 946
Reward: $0.06
Qualifications:

• GoodAtATOMICdimensions Exists [ Request | How to Use ]
• PreviouslySubmittedEventAnnotations DoesNotExist [ Request | How to Use ]
• Total approved HITs GreaterThanOrEqualTo 1000,
• HIT approval rate (%) GreaterThanOrEqualTo 99,

[3H8TR1N9M3IWNDBK6IE233ZH0WXEC7]​

 

Ah, ok, not me then.

Though, I was up there in the 80% most of the time. Hm.

 

The Requester included this note:
Hi there! Thanks for doing such a great job with my HITs over the last few weeks!! I just gave you the qual for yet another round of classifying AI ending HITs, titled 'Pick the most sensible how-to guide or caption ending'. Thanks! --Rowan

No new qual hit, just a new qual

 

My feedback from 3 and 4 days ago, in backwards order:

Spoiler: stat spam

You've ranked the real ending as best or 2nd best 465/489 times, or 95%. Great job!
Ranking the real ending as the best is the main thing that I'm looking for. You ranked the real endings likely 96% of the time, unlikely 3% of the time, and gibberish 0%.
You ranked the AI-generated endings likely 7% of the time, unlikely 92% of the time, and gibberish 0%.

You've ranked the real ending as best or 2nd best 220/223 times, or 98%. Great job!
Ranking the real ending as the best is the main thing that I'm looking for. You ranked the real endings likely 98% of the time, unlikely 1% of the time, and gibberish 0%.
You ranked the AI-generated endings likely 5% of the time, unlikely 94% of the time, and gibberish 0%.

You've ranked the real ending as best or 2nd best 44/47 times, or 93%. Great job!
Ranking the real ending as the best is the main thing that I'm looking for. You ranked the real endings likely 95% of the time, unlikely 4% of the time, and gibberish 0%.
You ranked the AI-generated endings likely 16% of the time, unlikely 82% of the time, and gibberish 0%.

You've ranked the real ending as best or 2nd best 15/16 times, or 93%. Great job!
Ranking the real ending as the best is the main thing that I'm looking for. You ranked the real endings likely 93% of the time, unlikely 6% of the time, and gibberish 0%.
You ranked the AI-generated endings likely 13% of the time, unlikely 86% of the time, and gibberish 0%.

You've ranked the real ending as best or 2nd best 153/156 times, or 98%. Great job!
Ranking the real ending as the best is the main thing that I'm looking for. You ranked the real endings likely 97% of the time, unlikely 2% of the time, and gibberish 0%.
You ranked the AI-generated endings likely 13% of the time, unlikely 85% of the time, and gibberish 0%.

My caption stats were pretty high as well. I got the '4' qual, but apparently not the first three. I'm curious if other people who worked on these at a high work quality level, like @jessers and @Calyrica got any of them. It's sort of weird to have what should be decent stats and not; the caption ones, where I think most folks did ~70%, I had for my last feedback:
You've ranked the real ending as best or 2nd best 80/88 times, or 90%. Great job!
Ranking the real ending as the best is the main thing that I'm looking for. You ranked the real endings likely 90% of the time, unlikely 9% of the time, and gibberish 0%.
You ranked the AI-generated endings likely 23% of the time, unlikely 75% of the time, and gibberish 0%.

I guess I'm curious what the heck the stats are of the people with the qual now, so I can reorient my perspective accordingly, as I put delicious pieces of bacon into my mac n cheese in bemusement.

 

Hah, sorry, I was jokingly worried cause he’s staring me down, hoping I give him more food

All the cats are good here at Club Birdie

 

I got 94% + on the most recent type, and 69-74% on the first type. I didn't get this new qual. I'm wondering how he picked people for these newest ones, 95%+ on both sets?