# Sample Variability & Sample Size

 Sampling & Sample variability Review of calculate Sample Statistics & making Box plots Sample Variability - how each sample varies from other samples (but is similar) iNZight VIT on sampling demonstration Sampling error notes (link to mathstatsfacilitators website) To Achieve you MUST discuss sample variability, including the variability of the population estimates. An understanding relating to sampling variability and variability of estimates must be evident. Another sample will give different medians and informal confidence intervals. Video Chris Wilds overview of sample variation, bootstrapping & randomisation Add in iNZight VIT on sample variation Qualifications & Work Data Cards: Data set csv Variables | sets of axis population boxplots & dot plots One data card | All data cards | excel version | Sample Size Activity: take sample from a population (data cards) manipulate, make box plots, compare with rest of class, increase sample size The effect of Sample size: on sample variability, box plots and representation of the population Samples of size n = 11 compared with n= 31 What sample size is most apppropriate? Increasing the sample size increases the cost of the survey. As long as the sampling method is sound there is very little benefit in increasing the sample size above 35. We would get little extra useful information. Note less variability in the medians: increasing sample size improves estimate. Draw in the interval for the sample medians. Note less variability in the inter-quartile range Note that estimates vary, even with larger sample size, and that it would be better to provide a range of possible values for the parameter, based on our estimate, rather than stating one value. Class notes, Blank notes Video Chris Wilds overview of sample variation, bootstrapping & randomisation Links to Census at school Animations page Sample size n = 10 animation Sample size n = 100 animation Sample size n = 1000 animation Key Concepts
Taking a sample
Calculate Sample Statistics
Box plots
Sample Variability

A population has (usually) unknown
parameters: mean and median

Take a sample eg size n = 9

The sampling process should ensure the sample represents the population
A biased or poor sample will give poor population estimates

Calculate sample statistics

Make a box plot

Compare your box plot with other samples

What might the population median be?

We could use out sample median as an estimate, but we have a problem… ##### Sample:

About Box plots (Murdoch university)

Max =

Upper Quartile =

Median =

Lower Quartile =

Minimum =

# ### Sample size n = 1000 animation   