Class Notes (1,016,082)

CA (584,127)

York (41,482)

Kinesiology & Health Science (2,420)

KINE 2049 (75)

Merv Mosher (71)

Lecture 8

# KINE 2049 Lecture Notes - Lecture 8: Multistage Sampling, Cluster Sampling, Systematic SamplingPremium

3 pages27 viewsFall 2017

School

York UniversityDepartment

Kinesiology & Health ScienceCourse Code

KINE 2049Professor

Merv MosherLecture

8This

**preview**shows half of the first page. to view the full**3 pages of the document.**KINE 2049 F

October 16th

Lecture 8, Getting a sample

Having a bigger sample is usually better for getting more accurate data, but what really

matters is that it is representative of your population !

Samples should be more than 30 people, less than 50% of the population, and be

representative. If you can do 50% of the population, you might as well do all the

population !

Random selection is the best way to select for a representative population, and it has

two factors that must be followed !

•every person or subject that you’re dealing with has an equal chance of being

selected !

•the selection of one does not bias the chance of the others in being selected !

Selection types

simple random selection: putting everyones name in a hat. The problem with this is

that you need to be careful of the size of the paper you use. It has to be the same size

for everyone, regardless of the length of their name !

•With replacement: every time you select a name, you put it back in the hat !

•Without replacement: every time you select a name, you put it aside, not in the hat,

which makes the odds not the same for each person !

Stratiﬁed random selection: You divide the population into diﬀerent subgroups based

on characteristics or traits, like gender or religion. Say you have a group of 70%

women and 30% men. To sample this, you can take 7 women and 3 men, which is still

the same ratio of the population !

Systematic sampling: you use a system, like taking every 10th person from a group !

Cluster sampling: Assume you want to sample teachers in the Toronto area, and

there’s a 1000 schools. You do cluster sampling, which is taking the names of all the

schools in a hat, and randomly selecting 10 of them. You then survey every teacher in

the 10 schools that you sampled out of the 1000 !

Multistage sampling: assume each school has 100 teachers, so thats a 1000

interviews. You can break it down again by taking 10 teachers in each school, and

sampling them. So you have to do 100 interviews in total !

Non probability sampling examples: !

•deliberate sampling: like the Nurses health study where they only sampled nurses.

You select subjects with speciﬁc characteristics that you are studying !

•Convenience sampling: You pick subjects that are easy for you to get into contact

with !

###### You're Reading a Preview

Unlock to view full version

Subscribers Only

#### Loved by over 2.2 million students

Over 90% improved by at least one letter grade.