Each group cluster by itself is representative of the whole population Randomly

Each group cluster by itself is representative of the

This preview shows page 10 - 12 out of 16 pages.

Each group (cluster) by itself is representative of the whole populationRandomly pick one clusterEither use the picked cluster to be your sample or pick a sample from the cluster using a probability-based procedure• The clusters should be mutually exclusive and collectively exhaustive 群集应该是相互排斥的,并且是完整的– i.e., tutorials are mutually exclusive and cover everyone• The elements should be heterogeneous within a cluster, but homogeneous across clusters. 元素在群集中应该是异构的,但是全部群集应该是相同的– i.e., the composition of tutorials are very similar (e.g., in terms of academic, social-economic, demographic backgrounds)• Used in cases where the population is very dispersed or data collection cost to cover the whole population base is high. 用于人口非常分散或数据收集成本覆盖整个人口基数较高的情况。– There are 14 tutorials and collecting data from all 14 is very costlyExample: Select 20 students from class to survey their satisfaction with the unit(MKF2121)’s Moodle site• Students can be found at tutorials, but data collection takes a lot of effort• Are there clusters that we can identify in this case?• Each tutorial can be considered a cluster (need to define what each cluster is and what is your sampling frame)Heterogeneous within each tutorial (i.e., every tutorial includes students of different gender, level of academic preparedness, nationality, etc)Homogeneous (similar) across (tutorials are rather similar to each other)Don’t confuse stratified and cluster sampling• Stratified sampling builds on the concept of segmentation
Image of page 10
Segments (strata) are very different from each otherClusters are the opposite. Different clusters should be very similar• Composition of your sampleCluster sampling—your sample consists of elements from a single clusterStratified sampling—your sample consists of elements from all strataWhen to use non-probability sampling techniques?• Researcher is constrained in time or resources研究人员受到时间或资源的限制• For exploratory research• Target population is small or difficult to identify• Sampling error is a major concern抽样误差是一个主要问题2.3. Deciding on the sample sizeKey trade‐offs between large vs. small sample size• Small sample size: cheaper and faster to execute; but results are less reliable• Large sample size: more expensive, but more likely to be representative of the population and hence results are less prone to error.容易出错3.0 Ask yourself: is the sample thus generated “representative” of the target population? (did I miss any segment?)是这样生的样品代表”了人群吗?错过任何ExampleAlternative sampling plan that will yield a more representative sample• Sampling frame instruction – customers who patronize the casino during from 9am to 9pm over 3 days. 1 week day, and 2 weekend days (Saturday and Sunday). Why spread it
Image of page 11
Image of page 12

You've reached the end of your free preview.

Want to read all 16 pages?

  • Two '16
  • representative, secondary data

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

Stuck? We have tutors online 24/7 who can help you get unstuck.
A+ icon
Ask Expert Tutors You can ask You can ask You can ask (will expire )
Answers in as fast as 15 minutes