Sampling units are not easily identified!
Cluster sampling - sampling population is
i)divided into groups – clusters
ii)elements within each cluster are selected
Different levels of clustering are possible!
Cluster 1 – all provinces or territories
Cluster 2 – major cities
Cluster 3 – school boards
Cluster 4 – schools
This process is called multi-stage cluster sampling
NON-RANDOM (NON-PROBABILITY) SAMPLING DESIGNS
Does not follow the theory of probability!
Number of elements either unknown or cannot be individually identified
Types of non-random sampling designs
Consent to participate
-not based on an obvious or visible characteristic
-common among market research and media
Judgemental or purposive sampling
-Who can provide the best information?
-Application - construction of a historical reality or description of a phenomenon
-No need for information on sampling frame, sample elements, location....
-Inclusion of participants needed for the study
-Findings cannot be generalized
Few individuals are selected
Asked to identify other individuals
Cluster sampling - sampling population is i) ii) divided into groups clusters elements within each cluster are selected. cluster 1 all provinces or territories. Non-random (non-probabil i ty) sampl i ng designs. does not follow the theory of probability! number of elements either unknown or cannot be individually identified. consent to participate not based on an obvious or visible characteristic common among market research and media. Application - construction of a historical reality or description of a phenomenon. No need for information on sampling frame, sample elements, location characteristics of both random and non-random sampling designs. the greater the heterogeneity or diversity, the greater the number of respondents needed. 1: degree of accuracy level of variation. Objective: to estimate the age of students in this lecture hall: assume willing to tolerate an error of 0. 5 years, confidence level 95, assume standard deviation is within 2 years.