You're making me glad I missed the code and didn't submit it because I did that survey in like 5 minutes
$1.50 Survey on new login website for Starbucks' free Wi-Fi on Android devices https://www.mturk.com/mturk/preview?groupId=2NPDAT2D5NQYVHY943AGDTYPVHX73L
$1.00 Listen to English sentences and answer questions about them. https://www.mturk.com/mturk/preview?groupId=2WSQHOVR14IXHGFMA78EGWAFZC6764
Email I received about the HIT. I don't remember doing it that fast Thank you for contacting me regarding this study. In pretests of the study, the average completion time was approximately 10m. The average completion time by MTurk workers was 8m6s. We wanted to be generous in our acceptance/rejection rates, taking into account that some Turkers are indeed fast readers, so we set a threshold of 3m15s (or 195s) for HITs that could be accepted. Your completion time was under 195s (less than 3m15s), which is less than half (40%) of the average time that everyone else took to complete the study. A common practice in social science research is to discard data that are in the extremes, such as when people complete a task too quickly or too slowly as compared to the average or expected completion time for the task. We cannot be sure that these people paid full attention to the task, or responded in a non-random manner to the survey questions. Therefore, we cannot use their data. As you can see in the screenshot I have attached, I paid 557 people for their work and only rejected 28 workers for this HIT, which is small minority of workers (about 4%), and this is consistent with social science practice to discard about 5% of outliers in one’s dataset. You can rest assured that we took these decisions very seriously and based them on empirical data, and I am sorry if this caused you any inconvenience. Best, -- Q / Que-Lam Huynh, PhD
I swear Toto (my dog) seems to go super crazy during the Franklin HITs with the lowest audio. Come on!
Certainly a reasonable and well thought out response, Swig, can't say that I disagree with his judgement.
pretty concise explanation. That said, it seems harsh to reject the outliers, as someone always has to be an outlier. Edit, actually on second thoughts it makes sense why they reject the outliers (have taken too many psych surveys this morning, don't mind my internal debate). One thing that might have happened is you did not accept it right away, but took the survey, then came back to find it not accepted. I know I've done that a few times and try to remember to let the timer sit a while before submitting.
I agree with what they are saying but that was the last hit I did yesterday and I thought it took me 6 minutes because I laid my baby down for a nap while the clock was going. Oh well.
I think you should be paid for the hit and he should then remove your results form his data, that's ridiculous.
I don't have my iPad with me so I can't try this one but if you're at home and have an iPhone 4 or better and a iPad 2 or better (needs video camera) then this would be a good one https://www.mturk.com/mturk/preview?groupId=2KYZBAT2D5NQC1B5TFRJL5U4QOD26J