LIFE OR DEATH

[khalil]

SOMEONE PLEASE TELL ME WITHOUT USING THE WORD 'STRATA' THE DIFFERENCE BETWEEN STRATIFED AND STRATIFED RANDOM SAMPLE.
I JUST DONT GET IT!

[melaniej]

Do you mean stratified sample and stratified random sample?

[melaniej]

OK, I'll give it a shot. They are very similar concepts, and I doubt we will have to identify the difference between them in the exam.

Stratified sample is where the population is divided into separate sub groups (these are called strata) and then they select a separate sample from these subgroups, ensuring they appear in the same proportions as they do in the target population.

Stratified RANDOM sample: Is the same thing, except for the sample taken from these subgroups is done randomly, by obtaining a list of ALL members of the subgroup and then using a random sampling technique, such as random number generator.

So they are essentially the same thing, except for in random sampling, the stratification is done RANDOMLY, rather then being just chosen...

[khalil]

OMG thank u so much
i understand it now

[melaniej]

No worries.
1 tip: next time you make a thread, don't name it "LIFE OR DEATH" something simple like - "Help With Sampling Procedures" would suffice.

Good luck for next week!

[khalil]

BUT one question: in stratified sampling how do they select the sample if its not random?

[khalil]

haha i had to attract ppl somehow

[melaniej]

If it's in the psych thread, I'll be there...
I enjoy answering other people's questions. Feel it helps consolidate my knowledge!

[khalil]

well then ive got one more question:
Outline the strengths and limitations of random and stratified samples...please

[melaniej]

RANDOM SAMPLE:

Strengths:
- Easier to conduct than stratified (don't have to sort into strata)
- Everyone has an equal chance of being selected, and as a result, it is likely the sample will be representative.
Limitations:
- Samples can be biased (i.e. not everyone has an equal chance)


STRATIFIED SAMPLE:

Strengths:
- Ensures all subgroups of a population are represented in the study
- Same proportions gives most accurate results.
Limitations:
- Can be exceptionally time consuming to divide up all the subgroups, and randomly sample.

[rh]

and in addition to those...

stratified sampling:
It’s advantages are that it allows a specific strata to be studied and compared with the rest of society and that it makes the sample representative on a key criteria. (eg/ age, gender)
It’s disadvantages are that access to specialist groups may be difficult, cause bias in the way that participants are selected and any generalisations cannot be necessarily applied to the whole population and that it's time consuming to administer.

random sampling:
It’s advantages are that it eliminates bias in the way that participants are chosen and any possible generalisations made are easier to apply to the whole population.
It’s disadvantages are that generalisations can’t be applied to specialist groups. (eg/ just women, widows, people with a mental disorder.)

[arthurk]

time is money so sometimes they say that it uses up resources as well

[Amnesiac]

Precisely