Numeracy & Quantitative Methods

Laura Lake

Types of Sampling
Probability sample a method of sampling that uses of random selection so that all units/ cases in the population have an equal probability of being chosen. Non-probability sample does not involve random selection and methods are not based on the rationale of probability theory.

Probability Sampling in Quantitative Research

Why is probability sampling important in quantitative research?

Research finding not based on samples that are biased / unrepresentative.

Based on a sampling frame it enables research to be replicable or repeatable. Research results can be projected from the sample to the larger population with known levels of certainty/precision (i.e. standard
errors & confidence intervals for survey estimates can be constructed).

Probability Sampling in Quantitative Research

To achieve this the sampling frame used needs to:
ensure that the correct population is being sampled i.e. it addresses the questions of interest

accurately covers all members of the population being studied so they have a chance to be sampled.

The quality of the population list (sampling frame) i.e. whether it is up-to-date and complete is the most important feature for accuracy in the sampling.

Types of Probability Sampling

Four main types of probability sampling:

1.Simple random sample

2.Systematic sample 3.Stratified random sample 4.Cluster/ multi-stage random sample

Simple Random Sampling

Randomly selecting units from a sampling frame.
Random means mathematically each unit from the sampling frame has an equal probability of being included in the sample.
Stages in random sampling:
Define population Develop sampling frame Assign each unit a number Randomly select the required amount of random numbers Systematically select random numbers until it meets the sample size requirements

Systematic Sampling
Similar to simple random sample.
No table of random numbers select directly from sampling frame.
Select according to fraction (100
sample from 1,000 frame then 10% so every 10th unit)

Define population

Develop sampling frame

Decide the sample size

Work out what fraction of the frame the sample size represents

First unit select by random numbers then every nth unit selected
(e.g. every 10th)

Stratified Random Sample

Gold standard of sampling. Why? Designed to be more representative of the population where the sampling frame is stratified according to population variables . Variables selected for stratifying are determined by the characteristics needed by the research. Stratification splitting the population into the different strata (variables e.g. gender, age, ethnic background). Samples can be stratified across more than one variable.

Stratified Random Sample

As a random sample:

Define population

Develop sampling frame according to characteristics required

Determine the proportion of each population variable of interest

Systematic sampling methods can then be followed to select sample unit

Cluster/ multi-stage random sample

Cluster sampling: selecting a sample based on specific, naturally occurring groups (clusters) within a population. - Example: randomly selecting 20 hospitals from a list of all
hospitals in England.
Multi-stage sampling: cluster sampling repeated at a number of levels. - Example: randomly selecting hospitals by county and then a sample of patients from each selected hospital.

Non-Probability Sampling
Three main types of non-probability sampling: 1.Convenience

2.Quota
3.Snowball

Convenience Sampling
A sample selected for ease of access, immediately known population group.

+ good response rate.

cannot generalise findings (do not know what population group the sample is representative of) so cannot move beyond describing the sample.

Quota Sampling
Aim is to sample reflecting proportions of population in different categories or quotas (e.g. gender, age, ethnicity). Used in often in market and opinion poll research. + easy to manage, quick

only reflects population in terms of the quota, possibility of bias in selection, no standard error

Snowball Sampling
Useful when a population is hidden or difficult to gain access to.

The contact with an initial group is used to make contact with others.
+ access to difficult to reach populations (other methods may not yield any results). - not representative of the population and will result in a biased sample as it is self-selecting.

Sample Size?
How large should my sample be in order for it to be representative? Larger samples are not necessarily better how representative a sample it depends on the sampling technique used and the size of the population. Determining sample size is dependent of how much error you are prepared to accept in your sample.

Sampling Error and Confidence

The larger the sample size the more likely error in the sample will decrease.
But, beyond a certain point increasing sample size does not provide large reductions in sampling error. Accuracy is a reflection of the sampling error and confidence level of the data.

Sampling Error and Confidence

If a sample has been selected according to probability we can assess the level of confidence. Confidence levels will allow you to state, with a certain level of confidence, that the sample findings would also be found in the population.

Confidence Intervals
Example:
Voting behaviour Labour Conservative Liberal Democrat Other % of poll 37% 35% 22% 6%

+/ - 3% at 95% confidence level

A confidence interval of +/- 3% at the 95% confidence level means that, 95% of the time, the true answer will be within 3% of the survey findings.

References
Bryman, A. (2008) Social Research Methods. 3rd Ed. Oxford: Oxford University Press. David, M. and Sutton, C. (2004) Social Research :The Basics. London: Sage. ESRC Survey Measurement Programme. Online: available from Survey Resource Network http://www.surveynet.ac.uk/ Oppenheim, A. (2000) Questionnaire Design, Interviewing and Attitude Measurement. London: Continuum

This resource was created by the University of Plymouth, Learning from WOeRk project. This project is funded by HEFCE as part of the HEA/JISC OER release programme. This resource is licensed under the terms of the Attribution-Non-Commercial-Share Alike 2.0 UK: England & Wales license (http://creativecommons.org/licenses/by-nc-sa/2.0/uk/). The resource, where specified below, contains other 3rd party materials under their own licenses. The licenses and attributions are outlined below:
1. 2. The name of the University of Plymouth and its logos are unregistered trade marks of the University. The University reserves all rights to these items beyond their inclusion in these CC resources. The JISC logo, the and the logo of the Higher Education Academy are licensed under the terms of the Creative Commons Attribution -non-commercial-No Derivative Works 2.0 UK England & Wales license. All reproductions must comply with the terms of that license.
Author
Institute Title Description Date Created Educational Level
Laura Lake

University of Plymouth
Numeracy & Quantitative Methods Sampling: Probability & non-probability sampling Overview of probability and non-probability sampling techniques in quantitative research. March 2011. Level 5 UKOER LFWOERK UOPCPDRM Learning from Woerk WBL Work Based Learning CPD Continuous Professional Development Probability sample, non-probability sample, simple random sample, systematic sample, stratified random sample, cluster/ multi-stage random sample, stratification, convenience sampling, quota sampling, snowball sampling, sampling error, confidence intervals. Attribution-Non-Commercial-Share Alike 2.0 UK: England & Wales license

Keywords
Creative Commons License

University of Plymouth, 2010, some rights reserved

Back page originally developed by the OER phase 1 C-Change project