Scribd
Upload
22 views

Blocking and Confounding in The 2 Factorial Design

Blocking is a technique used in experiments to account for background effects that cannot be held fixed but should not be left to vary randomly. It involves subdividing the experimental runs into equal blocks so that within each block, background effects are fixed, though they may vary between blocks. The experimental factor levels are then randomized within each block. Blocking helps reduce random error and ensure background effects do not interact with the experimental factors under study. The document then transitions to discussing confounding in 2k factorial designs.

Uploaded by

fawzi5111963_7872830
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
22 views

Blocking and Confounding in The 2 Factorial Design

Blocking is a technique used in experiments to account for background effects that cannot be held fixed but should not be left to vary randomly. It involves subdividing the experimental runs into equal blocks so that within each block, background effects are fixed, though they may vary between blocks. The experimental factor levels are then randomized within each block. Blocking helps reduce random error and ensure background effects do not interact with the experimental factors under study. The document then transitions to discussing confounding in 2k factorial designs.

Uploaded by

fawzi5111963_7872830
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 16

Blocking and

Confounding in The 2k
Factorial Design

Blocking Two-level Designs


Blocking
Example
Four

Blocks
Exercise

Blocking

In most experiments, causal factors for y (be they


known or unknown) are subdivided into three groups:

1) Factors whose effects are under study, by systematically


varying their levels according to an experimental design's
specifications ("experimental factors");
2) Factors not under study, whose levels are held fixed for the
entire experiment ("background effects");
3) Factors whose levels and effects on y are uncontrolled
("random error").

Note that for any experiment to be really useful, the


"background effect" factors must not interact with
the factors under study.
Also, the factors with really BIG effects should not be
left to vary out of control as random error.

Blocking
In

many instances factors which are not of


interest cannot be held fixed for the entire
experiment; yet, we do not want to let them
vary out of control. These are the instances
in which we can often use blocking.
Examples: Effects due to different

batches of raw material


operators,
machines,
time periods
Etc.

Blocking
When

we block an experiment, we
subdivide its runs into groups, usually of
equal size, called blocks. Ideally, within
each block the background effects are fixed.
It's o.k. if they vary block to block, without
interacting with the experimental factors.
The order of the runs within each block is
randomized. An additional Factor(s), Blocks,
is used in the analysis.

CONFOUNDING IN THE 2K Factorial Design

10

11

12

13

14

15

16

You might also like