Class Notes (1,100,000)

CA (620,000)

U of G (30,000)

SOAN (900)

SOAN 2120 (400)

David Walters (100)

Lecture

# SOAN 2120 Lecture Notes - Simple Random Sample, Nonprobability Sampling, Central Limit Theorem

by OC97365

School

University of GuelphDepartment

Sociology and AnthropologyCourse Code

SOAN 2120Professor

David WaltersThis

**preview**shows pages 1-2. to view the full**6 pages of the document.**November 27, 2012

Handing in the Final Assignment

- due outside of class Thursday (1-4pm)

- the late penalty is 5% per day

- after 5 pm on Thursday, it is LATE

- after the deadline, submit the assignment under the door (614 MACK)

- turnitin.com

o ID: 5632747

o Password: soan2120

- TOMORROW AFTERNOON (LAB): 20 MACK – 1 PM – 4PM

- Chapter 5 [ THIS SHIT IS IMPORTANT]

o Nonprobability sampling

Sample size in advance –they do not know the sample size in

advance

Random – they do not know if it is random

Knowledge of population – they know very little

Mathematical theory

Types of nonprobability samples (types, page 109)

o Probability sampling [EXACT OPPOSITE OF NONPROBABILITY]

Sample size in advance – must have this in advance

Random – must have

Knowledge of population

Mathematical theory – sampling methods are based heavily on

this

o Goal of Probability Sample?

To make

Populations, elements, and sampling frames

Population: boundaries

Geographical location

Businesses in Ontario

Employers at an organization

Students in the school

People who became parents I 2012

o Populations are abstract

Variable

Nevertheless, we need to estimate it

Generate list of what the population looks like

The list is the sampling frame (telephone directories, tax

records, driver’s license records, etc.)

Sample element: unit of analysis within the population

Sample frame should be a good representation of the

population

What if it’s not?

o Eg. telephone directories? Not everyone has a

telephone.

Only pages 1-2 are available for preview. Some parts have been intentionally blurred.

Sampling technique in order to collect sample

We want it to estimate the probability that everyone in

the population has a chance to be in our sample

SRS (simple random sampling)

Assign a number to each element in the sampling frame

Use a random number generator

Equal probability

Benefit of Random Sampling – sampling distribution

(clt) – inferences

Only with random sampling are the results we get from

our statistical test valid. The central limit theorem

applies only in the context that our sample was taken

randomly or through the weights….

Systematic sampling

First decide on your sample size

Population 1000

Sample 100

Sampling interval: 10

Start with your sampling frame (number each element):

pick one random number and then every 10 thereafter

Stratified Sampling

Divide population into sub-populations and then

sample within stratum

Stratum are picked by the researcher

o Eg. provinces, educational groups, religious

organizations, etc.

o So you stratify your population in these

groupings – this guarantees you get one of

each

Guarantees stratums are represented in the survey

(a certain number of people are chosen from each

stratum… like P.E.I., Inuit, Ph.D)

Weight variables allow us to adjust probability…

Cluster Sampling

Identify clusters (i.e. cities, blocks, households)

o Simple random sample clusters

o Simple random sample units within clusters

Population Guelph

o 1: City blocks

o 2: households (within the blocks)

o 3: individuals (within the households)

implications: we get three levels of sources of

random error

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

Unlock to view full version