Questions about HITs

[JaneDoe]

I apologize in advance if this is in the wrong section.

I'm trying to find out how to Turk quickly while avoiding rejections. Some of these might seem like dumb questions on the surface, but I'm curious to know the full extent of each.

Here are my questions:

1. Do I need to be honest while doing surveys?
2. Is it ok to complete surveys a few minutes before the estimated time?
3. What happens if I do the same survey twice, but the second one a week after the first?
4. If I accept & submit a HIT that I'm not fully qualified for, what are the chances of it getting rejected?
5. Is it true that it's best to take out the earnings only after you reach $20?
6. Does the $0.00 / 20.00 affect anything if it goes higher than 20?
7. Isn't there more risk involved when doing batches?
8. What do I do when a requester doesn't pay the bonus they promised?

It's better to be prepared.

I'll add more questions when something comes up. :o If you have any more suggestions on how to prevent rejections, I'm all ears.

Domo Arigato~

 

[jan]

I apologize in advance if this is in the wrong section.

I'm trying to find out how to Turk quickly while avoiding rejections. Some of these might seem like dumb questions on the surface, but I'm curious to know the full extent of each.

Here are my questions:

1. Do I need to be honest while doing surveys?
2. Is it ok to complete surveys a few minutes before the estimated time?
3. What happens if I do the same survey twice, but the second one a week after the first?
4. If I accept & submit a HIT that I'm not fully qualified for, what are the chances of it getting rejected?
5. Is it true that it's best to take out the earnings only after you reach $20?
6. Does the $0.00 / 20.00 affect anything if it goes higher than 20?
7. Isn't there more risk involved when doing batches?
8. What do I do when a requester doesn't pay the bonus they promised?

It's better to be prepared.

I'll add more questions when something comes up. If you have any more suggestions on how to prevent rejections, I'm all ears.

Domo Arigato~

1. Do I need to be honest while doing surveys? Yes. It's really important. Researchers trust turkers - it's how we get work. They do studies on this and have found that we are reliable. We don't want to jeopardize their research with invalid data (or jeopardize our income - who would survey a group of people who are not trustworthy?
   
2. Is it ok to complete surveys a few minutes before the estimated time? Sure. It's good to know that researchers have, at times, rejected for going too fast. But the times they give out are averages. Some people are slightly ahead of the average, others somewhat behind. You need to take the survey responsibly.
   
3. What happens if I do the same survey twice, but the second one a week after the first? You should not take a survey twice, unless the researcher says that it is okay. The default is not to do this.
   
4. If I accept & submit a HIT that I'm not fully qualified for, what are the chances of it getting rejected? I hope very good.
   
5. Is it true that it's best to take out the earnings only after you reach $20? I take my money out daily.
   
6. Does the $0.00 / 20.00 affect anything if it goes higher than 20? Not sure what you are asking.
   
7. Isn't there more risk involved when doing batches? Yes, a risk of a mass rejection. Read TO and do your best to do what the requester is asking.
   
8. What do I do when a requester doesn't pay the bonus they promised? Email them and ask what happened. If they don't respond, you could email the IRB and ask.

 

[kryss]

• Do I need to be honest while doing surveys?

Yes, it's unethical to be dishonest on surveys.

• Is it ok to complete surveys a few minutes before the estimated time?

Yes

• What happens if I do the same survey twice, but the second one a week after the first?

It's unethical to do the same survey twice, unless the requester specifies that it's okay to complete again.

• If I accept & submit a HIT that I'm not fully qualified for, what are the chances of it getting rejected?

It is unethical to do this and MTC does not support unethical behavior on mturk.

• Is it true that it's best to take out the earnings only after you reach $20?

Take out earnings as often as you like, although most people don't like leaving money as you run the risk of losing it if your account is suspended. (you can only transfer out once a day)

• Does the $0.00 / 20.00 affect anything if it goes higher than 20?

No

• Isn't there more risk involved when doing batches?

You should never do more hits for one requester that could tank your approval rating under 98%. It depends on the requester, if they have a history of rejecting, if they're new, etc. If you do quality work, you shouldn't have anything to worry about.

1. What do I do when a requester doesn't pay the bonus they promised?

Email

 

[JaneDoe]

Hmm, thank you for your replies. I don't mean to start a debate on what's ethical but there are a few nuisances. I feel as though you are replying in a way to not break the rules of your own site. For example:

1. Do I need to be honest while doing surveys? Yes. It's really important. Researchers trust turkers - it's how we get work. They do studies on this and have found that we are reliable. We don't want to jeopardize their research with invalid data (or jeopardize our income - who would survey a group of people who are not trustworthy?

But there have been members on MTC who did surveys that had nothing to do with them (i.e, targeted towards a certain gender [they did both]), no? How does MTC or mTurk distinguish between who is lying and who is not?

I don't believe that people are 100% honest on every survey. Otherwise, how would they finish HITs so quickly? I notice that I pay attention to every detail of a survey even though it slows down my PE significantly. I'm wondering if I should just start speeding through HITs, but at the same time I don't know what that'll do to my approval rate.

I'm trying to figure out a way to improve my daily Total Amount.

 

[jan]

Hmm, thank you for your replies. I don't mean to start a debate on what's ethical but there are a few nuisances. I feel as though you are replying in a way to not break the rules of your own site. For example:



But there have been members on MTC who did surveys that had nothing to do with them (i.e, targeted towards a certain gender [they did both]), no? How does MTC or mTurk distinguish between who is lying and who is not?

I don't believe that people are 100% honest on every survey. Otherwise, how would they finish HITs so quickly? I notice that I pay attention to every detail of a survey even though it slows down my PE significantly. I'm wondering if I should just start speeding through HITs, but at the same time I don't know what that'll do to my approval rate.

I'm trying to figure out a way to improve my daily Total Amount.

I wasn't picking my words for the site - it's really what I do and what I believe. We have (and most sites have, btw) looked at the data that supports the use of mturk. It's a source of income for us, but it's a tool for researchers. They only do research with mturk because they can validate their research. It really is based on us being a reliable population for study.

I'm not sure what else to say besides that. I really believe it. If you are looking for tips on improving your income, join in the main thread and follow along. Also, develop your skills at batch work. Learn some simple coding. That helps a great deal.

 

[JaneDoe]

I wasn't picking my words for the site - it's really what I do and what I believe. We have (and most sites have, btw) looked at the data that supports the use of mturk. It's a source of income for us, but it's a tool for researchers. They only do research with mturk because they can validate their research. It really is based on us being a reliable population for study.

I'm not sure what else to say besides that. I really believe it. If you are looking for tips on improving your income, join in the main thread and follow along. Also, develop your skills at batch work. Learn some simple coding. That helps a great deal.

I appreciate your honesty. I guess I'll just have to get better at Turking, because I spend far too much time pondering a question, which slows down my overall progress. My answers often depend on my mood and I don't have very strong opinions on a great majority of topics. Like for politics, I usually put "independent" or "moderate" because I don't care one way or the other. Sometimes I'll put more liberal or conservative just to appease the requester, lol. For me it's not really a question of ethics, or even principles, at least in regards to everyday socio-economic factors.

I have strong moral principles but they seldom come up in these surveys. Social constructs are constantly shifting and the way people view right from wrong is not what it once was. I keep these things in mind while I pick my answers. I realize that most likely misunderstand me and would soon label me an extremist of some kind. The best I could hope for is to answer as closely as possible even though many questions asked are quite vague.

Context is important to me...

 

[electrolyte]

Hmm, thank you for your replies. I don't mean to start a debate on what's ethical but there are a few nuisances. I feel as though you are replying in a way to not break the rules of your own site. For example:



But there have been members on MTC who did surveys that had nothing to do with them (i.e, targeted towards a certain gender [they did both]), no? How does MTC or mTurk distinguish between who is lying and who is not?

I don't believe that people are 100% honest on every survey. Otherwise, how would they finish HITs so quickly? I notice that I pay attention to every detail of a survey even though it slows down my PE significantly. I'm wondering if I should just start speeding through HITs, but at the same time I don't know what that'll do to my approval rate.

I'm trying to figure out a way to improve my daily Total Amount.

Don't get confused between workers sharing surveys here that have nothing to do with them and workers doing surveys that have nothing to do with them. I and others often share HITs that we don't qualify for because others will qualify. Just because I can't do it doesn't mean I can't or won't post the HIT here anyway to let others see it and check it out.

Requesters often over estimate the time we need to do surveys because they're giving generous time allowance or think we're reading their questions for the first time. Almost all surveys use some variation of the same measures which long time workers have seen dozens or hundreds of times before. I can answer some of those by rote memory because I literally have the questions memorized and know which answers are true for me. It doesn't take me nearly as much time as a requester thinks it might if they're assuming I've never encountered the question before. Newer workers here are certainly going to take longer since the questions are newer but for long time workers, we basically have the question sets memorized and each survey is just a different rearrangement of frequently used question sets with a few other questions thrown in. This is how we do surveys faster than the requester's estimated times. I would not recommend speeding through anything at the cost of possibly giving poor quality work. However, as you do more surveys, you will naturally get faster.

 

[electrolyte]

Like for politics, I usually put "independent" or "moderate" because I don't care one way or the other. Sometimes I'll put more liberal or conservative just to appease the requester, lol.

I would strongly discourage this. If requesters are getting responses from MTurk that just fit what you think they want and aren't valid responses (and they oftentimes can detect this because they or others have run the experiments before and/or have lab results to compare to), they won't continue using MTurk. That means less work here for all of us.

 

[JaneDoe]

I would strongly discourage this. If requesters are getting responses from MTurk that just fit what you think they want and aren't valid responses (and they oftentimes can detect this because they or others have run the experiments before and/or have lab results to compare to), they won't continue using MTurk. That means less work here for all of us.

I'll keep that in mind. I wasn't aware that it affects anyone and not just myself.