blocks: A set of 100 plastic "blocks"
In rwoldford/loon.data: Data Used to Illustrate 'Loon' Functionality

Description Format Author(s) See Also

A population of 100 "blocks" of different shapes and sizes cut from a single sheet of opaque plastic a few millimetres thick.

The blocks are of uniform thickness and density (all blocks were cut from the same opaque plastic sheet of about 5 mm thickness), but have different shapes. Each block has a number from 1 to 100 etched on one side. Ambiguous block numbers (e.g. 6 and 9) were disambiguated by placing a decimal place at the bottom of the number (e.g. 6. and 9.).

The blocks are treated as an entire population of 100 and presented in their entirety, spatially mixed, with the number side up. Students are asked to visually judge the blocks and to select 10 blocks whose average weight matches the average weight of all 100. Each student independently selects 10 block ids to serve as their judgment sample (see judgment).

The group variable identifies two strata ("A" or "B") roughly determined by the apparent area/volume of the blocks.

The data can be used to illustrate different sampling methods (e.g. simple random sampling, stratified sampling, and regression estimates based on modelling weight as a function of perimeter). Histograms of resulting estimates (over several samples) can be revealing.

A data frame with 100 rows and 4 variates

id: The id number etched on the block.
weight: The block's weight in grams.
perimeter: The perimeter length of the block in centimetres.
group: Group identification for the block: A are smaller blocks, B are larger.

R. Jock Mackay and R. Wayne Oldford

judgment

rwoldford/loon.data documentation built on May 17, 2021, 7:45 a.m.