Nested and Split-Plot Designs

• These are multifactor experiments that have

some important industrial applications.
• Nested and split-plot designs frequently
involve one or more random factors.
• There are many variations of these designs –
for example two stage, three stage nested
designs, split plot, split-split plot designs.

1
Nested Design

DOX 5E Montgomery 2
 Two-Stage Nested Design
• In a nested design, the levels of one factor (B) is
similar to but not identical to each other at
different levels of another factor (A)
• Consider a company that purchases material from
three suppliers
– The material comes in batches
– Is the purity of the material uniform?
• Experimental design
– Select four batches at random from each supplier
– Make three purity determinations from each batch

DOX 5E Montgomery 3
Two-Stage Nested Design

DOX 5E Montgomery 4
 Two-Stage Nested Design
Statistical Model and ANOVA
 i 1, 2,..., a

yijk    i   j (i )   (ij ) k  j 1, 2,..., b
k 1, 2,..., n

SST SS A  SS B ( A)  SS E
df : abn  1 a  1  a (b  1)  ab(n  1)

DOX 5E Montgomery 5
 Two-Stage Nested Design
Example 14-1

Three suppliers, four batches (selected randomly) from

each supplier, three samples of material taken (at
random) from each batch
Experiment and data, Table 14-3
Data is coded
Minitab balanced ANOVA will analyze nested designs
Mixed model, assume restricted form

DOX 5E Montgomery 6
DATA

DOX 5E Montgomery 7
MINITAB STEPS

8
MINITAB MODEL SETTING

9
 Minitab Analysis
Factor Type Levels Values
Supplier fixed 3 1 2 3
Batch(Supplier) random 4 1 2 3 4
Analysis of Variance for purity

Source DF SS MS F P
Supplier 2 15.056 7.528 0.97 0.416
Batch(Supplier) 9 69.917 7.769 2.94 0.017
Error 24 63.333 2.639
Total 35 148.306

Source Variance Error Expected Mean Square for Each Term

component term (using restricted model)
1 Supplier 2 (3) + 3(2) + 12Q[1]
2 Batch(Supplier) 1.710 3 (3) + 3(2)
3 Error 2.639 (3)
DOX 5E Montgomery 10
 Practical Interpretation – Example 14-1
• There is no difference in purity among suppliers, but
significant difference in purity among batches
(within suppliers)
• What are the practical implications of this
conclusion?
• Examine residual plots – pg. 562 – plot of residuals
versus supplier is very important (why?)
• What if we had incorrectly analyzed this experiment
as a factorial? (see Table 13-5, pg. 561)
• Estimation of variance components (pg. 565)

DOX 5E Montgomery 11
 Variations of the Nested Design
• Staggered nested designs (Pg. 566)
– Prevents too many degrees of freedom from building up
at lower levels
– Can be analyzed in Minitab (General Linear Model) – see
the supplemental text material for an example
• Several levels of nesting (pg. 566)
– The alloy formulation example
– This experiment has three stages of nesting
• Experiments with both nested and “crossed” or
factorial factors (pg. 569)

DOX 5E Montgomery 12
THREE STAGE NESTED DESIGN

13
THREE STAGE NESTED DESIGN
LAYOUT

14
Three-Stage Nested Design
Statistical Model

15
Three-Stage Nested Design
ANOVA (All fixed factors)

DOX 5E Montgomery 16
 Three-Stage Nested Design
ANOVA (Both Fixed & Random factors)

DOX 5E Montgomery 17
 Three-Stage Nested Design
Example (Q14.5 from Book)
• Consider the three-stage nested design
shown in Figure 14.5 to investigate alloy
hardness. Using the data that follow,
analyze the design, assuming that alloy
chemistry and heats are fixed factors and
ingots are random. Use the restricted form
of the mixed model.

DOX 5E Montgomery 18
DATA (Q14.5 from Book)

DOX 5E Montgomery 19
DOX 5E Montgomery 20
DOX 5E Montgomery 21
MINITAB OUTPUT

22
 Experiments with both nested
and “crossed” or factorial factors

23
Mixed Nested and Factorial Design
Statistical Model

24
Example 14-2 Nested and Factorial Factors

 i 1, 2,3
 j 1, 2

yijkl    i   j  k ( j )  ( )ij  ( )ik ( j )   (ijk )l 
 j 1, 2,3, 4
 l 1, 2
DOX 5E Montgomery 25
 Expected Mean Squares

Assume that fixtures and layouts are fixed, operators are

random – gives a mixed model (use restricted form)

DOX 5E Montgomery 26
MINITAB STEPS

27
MINITAB STEPS

28
Minitab Analysis

29
 The Split-Plot Design
• The split-plot is a multifactor experiment where it
is not possible to completely randomize the order
of the runs

30
DOX 5E Montgomery 31
DOX 5E Montgomery 32
DOX 5E Montgomery 33
DOX 5E Montgomery 34
 The Split-Plot Design
Model and Statistical Analysis
yijk    i   j  ( )ij  k  ( )ik  (  ) jk
 i 1, 2,..., r

( )ijk   ijk  j 1, 2,..., a
 k 1, 2,..., b


35
Alternate Model for the Split-Plot
 i 1, 2,..., r

yijk    i   j  ( )ij  k  (  ) jk   ijk  j 1, 2,..., a
 k 1, 2,..., b


DOX 5E Montgomery 36
 The Split-Plot Design-Example
• Example – paper manufacturing
– Three pulp preparation methods
– Four different temperatures
– Each replicate requires 12 runs
– The experimenters want to use three replicates
– How many batches of pulp are required?

37
 The Split-Plot Design

• Pulp preparation methods is a hard-to-change

factor
• Consider an alternate experimental design:
– In replicate 1, select a pulp preparation method,
prepare a batch
– Divide the batch into four sections or samples, and
assign one of the temperature levels to each
– Repeat for each pulp preparation method
– Conduct replicates 2 and 3 similarly

DOX 5E Montgomery 38
 The Split-Plot Design
• Each replicate (sometimes called blocks) has been
divided into three parts, called the whole plots
• Pulp preparation methods is the whole plot treatment
• Each whole plot has been divided into four subplots
or split-plots
• Temperature is the subplot treatment
• Generally, the hard-to-change factor is assigned to the
whole plots
• This design requires only 9 batches of pulp (assuming
three replicates)

DOX 5E Montgomery 39
DATA

DOX 5E Montgomery 40
 Split-Plot ANOVA

Calculations follow a three-factor ANOVA with one replicate

Note the two different error structures; whole plot and subplot
DOX 5E Montgomery 41
 Data set
res PM T WP Replicates res PM T WP Replicates res PM T WP Replicates

30 1 1 1 1 28 1 1 4 2 31 1 1 7 3

35 1 2 1 1 32 1 2 4 2 37 1 2 7 3

37 1 3 1 1 40 1 3 4 2 41 1 3 7 3

36 1 4 1 1 41 1 4 4 2 40 1 4 7 3

34 2 1 2 1 31 2 1 5 2 35 2 1 8 3

41 2 2 2 1 36 2 2 5 2 40 2 2 8 3

38 2 3 2 1 42 2 3 5 2 39 2 3 8 3

42 2 4 2 1 40 2 4 5 2 44 2 4 8 3

29 3 1 3 1 31 3 1 6 2 32 3 1 9 3

26 3 2 3 1 30 3 2 6 2 34 3 2 9 3

33 3 3 3 1 32 3 3 6 2 39 3 3 9 3

36 3 4 3 1 40 3 4 6 2 45 3 4 9 3

DOX 5E Montgomery 42
Book Method I

DOX 5E Montgomery 43
Book Method I

DOX 5E Montgomery 44
Book Method II

DOX 5E Montgomery 45
Quality Engineering Method

DOX 5E Montgomery 46
Quality Engineering Method

DOX 5E Montgomery 47