 |
Analysis of Data from Designed Experiments |
 |
|
Diagnostics and Remedial Measures |
Example: An experiment was conducted at Agricultural Research
station , Bagalkot (Karnataka) to study the effect of crop
residue on the growth and yield of Rabi Sorghum. The design
is adopted for this is Randomized Complete Block Design with
6 treatments in 4 replications. The net plot size used is
4.20 ´
10.20 m2. The
treatment details are given below.
|
Treatment |
Treatment
Details |
|
1 |
Control |
|
2 |
50 kg/ha N as Urea + 25 kg/ha of P2O5
as Super Phosphate (Recommended dose) |
|
3 |
Subabul Stalks @ 5 Tons/ha |
|
4 |
Jowar Stubbles @ 5 Tons/ha
|
|
5 |
Jowar Stubbles @ 2.5 Tons/ha + Subabul Stalks @ 2.5 Tons/ha |
|
6 |
Jowar Stubbles @ 1.25 Tons/ha + Subabul Stalks @ 3.25
Tons/ha |
The
experimental data (yield in kg/plot) is
|
Treatments |
Replication |
|||
|
R1 |
R2 |
R3 |
R4 |
|
|
T1 |
8.90 |
9.30 |
9.20 |
9.10 |
|
T2 |
10.65 |
10.50 |
10.10 |
11.05 |
|
T3 |
10.60 |
11.25 |
10.25 |
11.50 |
|
T4 |
9.15 |
8.90 |
9.30 |
10.30 |
|
T5 |
13.00 |
9.91 |
10.14 |
8.85 |
|
T6 |
9.80 |
9.95 |
9.95 |
10.20 |
1.
Check
the assumptions of normality and homogeneity of error terms.
-
If errors are non-normal and / or heterogeneous, then apply Box-Cox transformation so as to make the errors from transformed data as normal and homogeneous.
-
If the errors remain non-normal and/or heterogeneous even after transformation, then apply Skilling and Macks non-parametric test for testing the equality of treatment effects.
Hint:
Fit a two-way classified, additive linear model and obtain
the residuals. Test the normality of errors using Kolmogorov-Smirnov
test or Shapiro-Wilk’s test. If errors are normally
distributed, then test the homogeneity of errors using
Home
Descriptive
Statistics
Tests of
Significance Correlation
and Regression Completely
Randomised Design
RCB
Design Incomplete
Block Design Resolvable
Block Design Augmented
Design Latin
Square Design Factorial
RCB Design Partially
Confounded Design Factorial
Experiment with Extra Treatments Split
Plot Design Strip
Plot Design Response
Surface Design Cross
Over Design Analysis
of Covariance Diagnostics
and Remedial Measures Principal
Component Analysis Cluster
Analysis Groups
of Experiments Non-Linear
Models
Copyright Disclaimer How to Quote this page Report Error Comments/suggestions
(Under Development)
For
exposure on SAS, SPSS, MINITAB, SYSTAT and
MS-EXCEL
for analysis of data from designed experiments:
Please see Module I of Electronic Book II: Advances in Data Analytical Techniques
available at Design Resources Server(www.iasri.res.in/design)