How to determine an appropriate sample size

[Janmiller]

Hi! I would appreciate if anyone of you can help me out here.

I wish to conduct a survey of the general population of a country and would like to determine the appropriate sample size. May I know how to do it?

Additionally, if I were to conduct a survey by approaching the people on the streets, will the survey become biased? Is this considered a simple random sampling method?

Thank you!

 

[OliverCurtis]

You have to consider how confident you want to be in your result. If you want 100% confidence then you have to ask everyone!

1. Eliminate members of the population that you know the survey is not going to be relevant to i.e. you don't ask those under 17 about their experiences learning how to drive or under 18's what their favourite beer is.
2. Take the remaining population and make a judgement on your confidence level.
3. If you ask everyone this is actually called a census.
4. Take into account your margin of error i.e. how many people will not tell the truth and other factors.

Unless you do the survey in a place that means that people who you will be asking will be obviously biased then no. For example if you were surveying people about public transport and you went on a day when a bus serving that area had broken down or in a place where there is particularly bad public transport then the answers you get may not be representative of the whole population.

 

[Dr.Peterson]

I wish to conduct a survey of the general population of a country and would like to determine the appropriate sample size. May I know how to do it?

Additionally, if I were to conduct a survey by approaching the people on the streets, will the survey become biased? Is this considered a simple random sampling method?

What have you been taught about choosing a sample size?

Do you know what kind of data you are gathering, what margin of error and confidence level you want, and what the formula is for the kind of data you have? Here's the first source Google gave me, which covers a number of cases: https://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_power/bs704_power_print.html

What have you been taught about what "simple random sampling" means? Do you know what kinds of sampling are biased?

This question suggests that this is for a course you are taking, not an actual survey; if it's from an actual problem you were given, please quote it as given to you, and tell us what topics have been taught.

 

[Janmiller]

You have to consider how confident you want to be in your result. If you want 100% confidence then you have to ask everyone!

1. Eliminate members of the population that you know the survey is not going to be relevant to i.e. you don't ask those under 17 about their experiences learning how to drive or under 18's what their favourite beer is.
2. Take the remaining population and make a judgement on your confidence level.
3. If you ask everyone this is actually called a census.
4. Take into account your margin of error i.e. how many people will not tell the truth and other factors.

Unless you do the survey in a place that means that people who you will be asking will be obviously biased then no. For example if you were surveying people about public transport and you went on a day when a bus serving that area had broken down or in a place where there is particularly bad public transport then the answers you get may not be representative of the whole population.

I see, now I got a better understanding of the sampling! Thanks! Do you think using the sample size formula is the right way to determine the sample size of a country population?

 

[Janmiller]

What have you been taught about choosing a sample size?

Do you know what kind of data you are gathering, what margin of error and confidence level you want, and what the formula is for the kind of data you have? Here's the first source Google gave me, which covers a number of cases: https://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_power/bs704_power_print.html

What have you been taught about what "simple random sampling" means? Do you know what kinds of sampling are biased?

This question suggests that this is for a course you are taking, not an actual survey; if it's from an actual problem you were given, please quote it as given to you, and tell us what topics have been taught.

Hi Dr.Peterson! Thanks for the response! There is no actual problem given to me, I am doing a project for school and need to determine the appropriate sample size for the Singapore population. I wish to find out how frequent Singaporeans had fast food. I have determined the margin of error as 5% and the confidence level as 95%, the population in Singapore currently is about 5.639 million. Hence, is using the sample size formula the right way to determine the sample size?

 

[Dr.Peterson]

Give it a try. Choose which formula to use (there are several in the link I gave, depending on what type of statistic you want (mean, proportion, ...), and the best way to gain an understanding of this is to try applying it.

Is this at least associated with a class you are taking in which you have learned about sampling? That's the main thing I was asking.

 

[Janmiller]

Give it a try. Choose which formula to use (there are several in the link I gave, depending on what type of statistic you want (mean, proportion, ...), and the best way to gain an understanding of this is to try applying it.

Is this at least associated with a class you are taking in which you have learned about sampling? That's the main thing I was asking.

Nope, which is why I am confused about how to go about doing this.

 

[tkhunny]

Do you know how to calculate a Confidence Interval? This holds the secret to determining sample size.

Meeting people on the street is NOT random. That doesn't mean it can't be done.