Anyone good at Stats? What sampling method is this?

[alex_boothby]

Hi, Anyone knowledgeable at stats. I hope you can help.

I collected my samples on the street by just asking passers by to fill in my questionnaire. I moved to 5 different locations within the city and spent 3 hours each visit 3 times a week for one month.

Is this random sampling? Or it it sill convenience sampling? Can anyone help me make it sound as good as possible! While i know this can never be representative, can argue that its more than convenience?

I also did ask every 3rd passer by.
Cheers!

[berry120]

From my understanding there's a fine line between the two - it's not truly random, but then you could argue it never is (as you point out.) If someone was really picking your sampling apart then they could say things like you've only sampled people within that city which isn't necessarily representative, you're more likely to sample people who venture outside often rather than sit at home all day, you're only sampling people that are willing to stop and talk - that sort of thing.

However, you could keep this up forever. What you have shown is that you've made a concerted effort to make it as representative as you realistically can (without knocking on doors or travelling huge distances) and for your purposes that's probably good enough (though you don't say what the purpose is?) As long as you identify the weaknesses in your approach such as I've outlined above, to show you're aware of those weaknesses, I'd say you can make the claim it's pretty representative.

[alex_boothby]

Thanks, that was pretty much confirming what I believed. It all seems a bit ... unclear. I suppose im just trying to get away from convenience sampling due to the stigma that comes with it, but i guess as long as I make effort to show how I did everything possible to make it fair as possible.

[ssc456]

I think it depends alot on what information your trying to gather . . . .

ie if your gatehring information of peoples jobs and your interview outside train stations between 4pm and 6pm every day then your likely to get a large group of commuters mainly in office jobs. However if you were asking them about what illness runs in their families then this could be considered irellivant.

However i also agree that you can never really have a random sample of people. As people (really annoying people) could argue that your sample of so called "office workers" could have more relatives who were "office workers" and therefore any illness associated with office work ie RSI or Arthritus or migrains etc would be biass.

But that depends on how far you dig down.

[alex_boothby]

I agree it can get very pernickety. I am trying to (w/o going into much detail) gather peoples opinion on climate related matters. And I want to see how education level effects opinion.

[IAntDemo]

I;m sorry what's stats?

[alex_boothby]

Statistics

[IAntDemo]

for what exactly? sampling? what do you mean.. sorry I just don't understand much of what you're saying.. I'm guessing it's music right?

[superman22x]

No music. Sampling, as in, taking samples of the public's opinion/knowledge on a subject.

[IAntDemo]

o, what subject lol