STRIP PLOT DESIGN Randomization and Layout: Model
In this design each block is divided into number The model for strip plot design is
Sometimes situation arises when two factors each of vertical and horizontal strips depending on the Yijk = i j ()ij k ()ik ()jk
requiring larger experimental units are to be
levels of the respective factors. Let A represent ijk i = 1,2, …,r, j = 1,2, …,a, k = 1,2,
tested in the same experiment, e.g., suppose four
levels of spacing and three levels of methods of the vertical factor with a levels, B represent the
ploughing are to be tested in the same horizontal factor with b levels and r represent the …b
experiment. Here both the factors require large number of replications. To layout the experiment, where,
experimental units. If the combination of the two the experimental area is divided into r blocks.
factors at all possible levels are allotted in a Yijk : observation
Each block is divided into b horizontal strips and
R.B.D. in the normal way, the experimental plots corresponding to kth
b treatments are randomly assigned to these strips
shall have to be very large thereby bringing level of factor A, jth
heterogeneity. So, it will not be appropriate. On in each of the r blocks separately and
level of factor B and ith
the other hand if one factor(spacing) is taken in independently. Then each block is divided into a
main plots and other factor(methods of
replication.
vertical strips and a treatments are randomly
ploughing) is taken in sub- plots within main : general mean
assigned to these strips in each of the r
plots, the sub-plots shall have to be large enough. i : ith block effect
blocks separately and independently. A possible
Hence split plot design also will not be j : effect of jth level of factor A
appropriate. In such situations a design called layout of a strip plot experiment with a 5 (a1,
k : effect of kth level of factor B
Strip plot design is adopted. a2, a3, a4, and a5) , b
()jk : interaction between jth level of factor A
3 (b1, b2, and b3) and four replications is given
The strip plot is a 2-factor design that allows for and the kth level of factor B
below:
greater precision in the measurement of the The error components ()ij, ()ik and ijk
interaction effect while sacrificing the degree of are independently and normally distributed
precision on the main effects. The experimental
with means zero and respective variances
area is divided into three plots, namely the
vertical-strip plot, the horizontal- strip plot, and 2a, 2b, and 2.
the intersection plot. We allocate factors A and B, Analysis
respectively, to the vertical and horizontal-strip In statistical analysis separate estimates
plots, and allow the intersection plot to The strip plot design sacrifices precision of error are obtained for main effects of
accommodate the interaction between these two on the main effects of both the factors in the factor, A and B and for their
factors. As in the split plot design, the vertical
order to provide higher precision on the interaction AB. Thus there will be three
and the horizontal plots are perpendicular to each
other. However, in the strip plot design the interaction which will generally be more error mean squares applicable for testing
relationship between the vertical and horizontal accurately determined than in either the significance of main effects of the
plot sizes is not as distinct as the main and sub- randomized blocks or simple split plot factors and their interaction separately.
plots were in the split plot design. The sub-plot design. Consequently this design is not Suppose 4 levels of spacings(A) and 3
treatments instead of being randomized recommended unless practical levels of methods(B) of ploughing are to
independently within each main plot as in the considerations necessitate its use or unless
case of split plot design are arranged in strips be tested in the same experiment. Each
across each replication. The intersection plot, the interaction is the principle object of replication is divided into 4 strips
which is one of the characteristics of the design, study. vertically and into 3 strips horizontally.
is the smallest in size. In the vertical strips the four different
levels of spacings are allotted randomly
and in the horizontal strips three
methods of ploughing are allotted
randomly.
Let there be 4 replications(R). The
analysis of variance is carried out in
three parts viz. vertical strip analysis,
horizontal strip analysis and interaction STRIP PLOT DESIGN
analysis as follows:
Form spacing replication (A R)
table of yield totals and from this table RANJAN KUMAR KAR
compute the
S.S. due to replication, S.S. due to spacings SASWATI SUBHALAXMI PATRA
and S.S. due to interaction - Replication
Spacing i.e. error(a).
Form method replication (B R) table
of yield totals and from this table
compute the
S.S. due to methods and S.S. due to
interaction - Replication Method i.e.
error(b).
Form spacing method (A B) table of
yield totals and from this table compute
the
S.S. due to interaction - Spacing
Method. 2025
Total S.S. will be obtained as usual
by considering all the observations
of the experiment and the error Standard Errors and Critical Differences:
S.S. i.e. error(c) will be obtained by
subtracting from total S.S. all the
S.S. for various sources. COLLEGE OF FORESTRY,
Now, calculate the mean square ODISHA UNIVERSITY OF
for each source of variation by
dividing each sum of squares by its
AGRICULTURE AND
respective degrees of freedom. TECHNOLOGY, BHUBANESWAR
Compute the F-value for each -751003
source of variation by dividing each
mean square by the corresponding
error term.
The analysis of variance table is outlined
as follows:
