GEOG 3MB3 Lecture Notes - Lecture 6: Cluster Sampling, Traffic Collision, Sampling Frame
September 27, 2016
Notation
- Use this notation when answering questions and when interpreting formulas he
provides you
Sampling defined
- A sample is a subset of a population
- Sampling is the choosing the subset of the population with some definable
method or procedure
- Informally the sample is just a collection of data with properties
- We use sampling because it is not practical to collect data from populations
Sampling Example
- We sample by picking names out of a hat and collect data in order to get an idea
of what is happening in a population
Sampling Unit
- Say we are trying to find how many people are involved with non-lethal motor
vehicle collisions
- We could sample based on persons by asking people if they were involved in a
collision
- Or you could look at car crashes at a given location – you could look at and
analyze police reports. But a better sampling unit would be what is happening at
the area over time so the sample unit could be time
Sampling unit cont’d
- Sampling unit for oil companies we can use oil companies as our sampling unit,
or you could just sample workers
- Average transportation…. The sampling unit here could be drawn from CMA
studies
- Infant… sampling unit here could be countries
Sample Frame
- Generally the bigger the population you draw from the larger the implications of
your research
- Generally the bigger your scope the bigger the population has to be
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Use this notation when answering questions and when interpreting formulas he provides you. A sample is a subset of a population. Sampling is the choosing the subset of the population with some de nable method or procedure. Informally the sample is just a collection of data with properties. We use sampling because it is not practical to collect data from populations. We sample by picking names out of a hat and collect data in order to get an idea of what is happening in a population. Say we are trying to nd how many people are involved with non-lethal motor vehicle collisions. We could sample based on persons by asking people if they were involved in a collision. Or you could look at car crashes at a given location you could look at and analyze police reports. But a better sampling unit would be what is happening at the area over time so the sample unit could be time.