Sampling Bias #2 by Helen Hailin Zhao

Question: if you conduct a survey on Mturk (an online crowdsourcing platform powered by Amazon) and recruit participants with $5 per survey to answer questions about their shopping preferences. What are the biases here?

hhzhao — 2024-09-11 23:59:20 UTC

hhzhao — 2024-09-12 00:00:12 UTC

3036342507 — 2024-09-13 11:10:51 UTC

voluntary response bias

Voluntary Response Bias

3036340494 — 2024-09-13 11:10:54 UTC

3036341723 — 2024-09-13 11:10:54 UTC

some customers don't buy things online

3036348848 — 2024-09-13 11:10:58 UTC

Voluntary Response Bias

3036342595 — 2024-09-13 11:10:59 UTC

The sample is not random. It's kind of like the voluntary response bias.

3036340406 — 2024-09-13 11:11:23 UTC

maybe the respondents would not share useful facts to the survey because they just want to get the 5 dollars

Voluntary response bias

3036341058 — 2024-09-13 11:11:23 UTC

Only people who are users of Mturk are involved in the survey.

2024-09-13 11:11:26 UTC

The reward maybe too little to some part of people,some young children and the elders aren’t willing to do such response which will also cause representation sample bias

wangsihan5600 — 2024-09-13 11:11:27 UTC

Some people maybe would not willing to talk about their preferences for getting 5 dollars.

3036342337 — 2024-09-13 11:11:35 UTC

This survey would only be responded by those who have time and can be motivated by $5.

3036347624 — 2024-09-13 11:11:37 UTC

People might answer the questions randomly just for the $5, which don’t truly represent their thoughts.

Voluntary response bias

3036342301 — 2024-09-13 11:11:38 UTC

People who choose to take the survey are probably more price sensitive, the volunteers may appear to be a typical group

3036166745 — 2024-09-13 11:11:42 UTC

Exclude the buyers who don't care about the $5.

3035638973 — 2024-09-13 11:11:42 UTC

conflicts of interest

3036342703 — 2024-09-13 11:11:44 UTC

Ignore other who are not willing to pay the fee. Voluntary bias

Voluntary Response Bias

3036341852 — 2024-09-13 11:11:52 UTC

Cause people with the strong idea, or interested in the money will be more likely to finish the survey.

voluntary response bias

3036341840 — 2024-09-13 11:11:53 UTC

People who won't be attracted by 5 dollars didn't be considered

3036343226 — 2024-09-13 11:11:53 UTC

voluntary responses bias and undercoverage

3036342557 — 2024-09-13 11:11:53 UTC

In order to reward quick answers or provide false information.

3036345468 — 2024-09-13 11:11:54 UTC

Voluntary response bias. They only select samples from one platform and therefore the participants are having certain characteristics of shopping preferences to use and getting used to the platform.

voluntary response bias

2024-09-13 11:11:57 UTC

where people are being incentives to do the questionnaire. The people tend to favor the platform that provide the monetary benefits (Amazon)

3036340951 — 2024-09-13 11:12:01 UTC

This survey only covers the people who think it's worth participating in this survey and earning $5.

3036340652 — 2024-09-13 11:12:03 UTC

There will be voluntary response bias because it is probable that only the ones who want the prize will participate the survey

3036348197 — 2024-09-13 11:12:03 UTC

people have higher income tend not to spend time in the survey

3036341450 — 2024-09-13 11:12:04 UTC

Only individuals willing to participate for compensation will respond

3036341280 — 2024-09-13 11:12:06 UTC

Conducting a survey on MTurk with paid participants introduces voluntary response bias. Because individuals who choose to participate are motivated by financial incentives, which may not reflect the broader population’s shopping preferences.

3036341242 — 2024-09-13 11:12:07 UTC

Offering $5 per survey might attract a specific subset of participants who are motivated primarily by monetary compensation rather than genuine interest in the survey topic. This could influence the quality and reliability of the responses.

3036342442 — 2024-09-13 11:12:07 UTC

There exists lots of people who are not willing to spend time on the survey for $5

3036342832 — 2024-09-13 11:12:08 UTC

Volunteers who do the survey for $5 return may be not the high-income shoppers who don’t care that return.

3036342765 — 2024-09-13 11:12:16 UTC

this is voluntary response bias because only those who care the 5$ would answer this survey.

3036348109 — 2024-09-13 11:12:16 UTC

part of the participants only want to get 5＄

3036342076 — 2024-09-13 11:12:17 UTC

The people answer the survey maybe are interested in the 5 dollars, not interested in the real survey.

3036348020 — 2024-09-13 11:12:19 UTC

This survey is more attractive for respondents who want to get 5$, so it shows a voluntary response bias.

3036341735 — 2024-09-13 11:12:20 UTC

These people may only want to get the 5 dollars but not answer the question with consideration, which may cause the survey containing many error statistic.

Voluntary response bias. The sampling is not convincing because the participants of the survey might not actually shop online.

3036342210 — 2024-09-13 11:12:21 UTC

3036341890 — 2024-09-13 11:12:21 UTC

Voluntary Response Bias. Only people who are motivated by $5 will respond.

3036343537 — 2024-09-13 11:12:22 UTC

There is voluntary response bias. The researcher could only collect data from people willing to share their shopping preferences. These people are quite different from people unwilling to participate in this survey.

3036341278 — 2024-09-13 11:12:24 UTC

some people answer the survey only in order to earn money
some people don't access to the survey and maybe even not interested in shopping online

Voluntary response bias

3036341668 — 2024-09-13 11:12:25 UTC

People who don't care about 5 dollars will not finish this survey, which means rich people do not sensitive to the bonus.

Voluntary bias, because people who did not use Mturk, their opinions cannot be included in this survey.

3036440755 — 2024-09-13 11:12:25 UTC

3036341371 — 2024-09-13 11:12:25 UTC

The users of Mturk may be not representative

Voluntary response bias. The participants are those people driven by the reward

3036341711 — 2024-09-13 11:12:25 UTC

Voluntary Response Bias

3036343484 — 2024-09-13 11:12:30 UTC

People who did the survey because of money have different purchasing habits with other people.

3036344189 — 2024-09-13 11:12:32 UTC

Voluntary bias. The group of people who really need the money will be more willing to take the survey, while for the group of people who don't care about the 5 dollar reward will not participate.

3036342715 — 2024-09-13 11:12:35 UTC

Using MTurk for surveys can introduce biases like self-selection and payment-induced response quality. Be cautious of non-representative samples and motivated biases due to incentives.

Still undercovereage biases

3036341955 — 2024-09-13 11:12:37 UTC

The sample would be limited to the customers that would scan Amazon so that they would be the only sample that can find that survey. The sampling is still not widely ranged.

3036340949 — 2024-09-13 11:12:42 UTC

The sampling only includes the participants who are using Mturk and those who are willing to do a survey for 5 dollars

2024-09-13 11:12:47 UTC

this is voluntary response bias, because many people answer the survey just for money

3036341541 — 2024-09-13 11:12:48 UTC

People who are willing to participate in paid surveys may differ from the general population. They may be more motivated to participate for the money, which can lead to responses that may be biased from those of the general population.

3036347856 — 2024-09-13 11:12:51 UTC

The demographic characteristics of MTurk workers may not be representative of the general population. For example, MTurk workers tend to be younger, more highly educated, and more technically savvy than the general population. This means that the results of the survey may not be applicable to all shoppers.

3036341670 — 2024-09-13 11:12:52 UTC

the shopping preferences of people who want to get 5 dollars may be different from those who don't need 5 dollars.

3036347777 — 2024-09-13 11:12:54 UTC

There’s a potential demographic bias since Mturk workers tend to have certain characteristics that may not reflect the general public.

3036341383 — 2024-09-13 11:12:55 UTC

Some people may just do the survey for the money and not answer their true thoughts

Voluntary Response Bias

3036342961 — 2024-09-13 11:12:59 UTC

Only people who are interested of $5 are involved in the survey.

Voluntary Response Bias

3036343410 — 2024-09-13 11:12:59 UTC

People who would like to take the money and do the survey are less likely to be of strong purchasing power and are more likely to go for cheap items.

3036341979 — 2024-09-13 11:13:00 UTC

Voluntary Response Bias, and there may be fewer samples with higher income levels

3036342753 — 2024-09-13 11:13:06 UTC

Workers may rush through surveys because they are paid to do so, rather than reading and thinking through the questions carefully, which can lead to lower quality data.

3637663 — 2024-09-13 11:13:07 UTC

Voluntary response biases

dddaijxuan — 2024-09-13 11:13:11 UTC

The range of the people are not random. Maybe they just want to acquire the fund and cannot provide the useful information.

3036441846 — 2024-09-13 11:13:15 UTC

ignore offline customer.

people may only want to get $5 but never not answer the survey in detail.

some people don't care about $5.

3036340614 — 2024-09-13 11:13:17 UTC

not all people want to do the surveys and the sample just cover the little groups

3036348654 — 2024-09-13 11:13:23 UTC

The experiment attracts participants who are more focused on monetary awards rather than the significance of the survey itself.

2024-09-13 11:13:24 UTC

Only people being interested in $5 will answer the questions. It’s voluntary response bias

3036160404 — 2024-09-13 11:13:25 UTC

the sample are probably from young generation who use the Internet

3036342624 — 2024-09-13 11:13:28 UTC

voluntary response bias-people will be more incentive to the survey for $5.

Voluntary response bias

3036340810 — 2024-09-13 11:13:35 UTC

The result may not be accurate because they recruit participants who are more willing to take part in than others. Participants just want to get $5 rather than giving their right answer.

3036343927 — 2024-09-13 11:13:35 UTC

People who are willing to be a paid participant would have something in common. They might be strongly motivated by money, which would affect their shopping behaviour.

voluntary response bias.the buyer who do not careabout the 5 dollars

2024-09-13 11:13:41 UTC

Voluntary Response Bias

3036343525 — 2024-09-13 11:13:44 UTC

The audience’s reaction of the shopping preferences mostly positive, because the people who answers the questions are usually more likely shopping online.

3036348537 — 2024-09-13 11:13:45 UTC

Voluntary Response Bias. People who attend this survey may just because of the rewards and lots of others from other platforms may not be involved.

3036341826 — 2024-09-13 11:13:49 UTC

Most people are users of Mturk and are motivated by $5, the result is trustless.

3036340705 — 2024-09-13 11:13:54 UTC

may get the random answer who just participate for $5

Voluntary response bias

3036342167 — 2024-09-13 11:13:58 UTC

The quality of the survey responses need to be worried since their target might only be the $5, and they may only want to answer as quick as possible.

3036159493 — 2024-09-13 11:14:00 UTC

people may not tell the truth in order to get 5 dollar

3036342090 — 2024-09-13 11:14:04 UTC

undercoverage: only Mturk's users will be included

conniexiao14 — 2024-09-13 11:14:06 UTC

voluntary response bias

Because people who decided to do the survey mostly do it for the money. And they represent a certain group of people.

3036341412 — 2024-09-13 11:14:15 UTC

The people who do the survey are those who are more willing to do online shopping, who prefer small money awards.

3036343551 — 2024-09-13 11:14:22 UTC

MTurk workers may not be representative of the general population. And the $5 reward could influence participants to participate for the sake of the money rather than a genuine interest in the survey topic, potentially affecting the authenticity of their responses.

Voluntary response bias

2024-09-13 11:14:27 UTC

Because who want to do it is different from what you want