You are on page 1of 7

AGRONOMY 112 - Problem Set 3

A. An experiment was conducted to compare yield performance of corn at four levels of


phosphorus using a split plot design in RCBD. The researcher is particularly interested in
the effects of phosphorus and used a design that would lend higher precision to t effects of
phosphorus. Three varieties of varying yields were used to verify if there is any interaction
of varieties with phosphorus.

Variety Block P0 P30 P60 P120


V1 1 80.25 90.9 91.2 89.4
V1 2 93.3 103.2 106.35 101.7
V1 3 80.1 89.25 91.5 90.45
V2 1 67.2 76.5 77.25 74.85
V2 2 78.75 88.05 89.1 87.15
V2 3 64.65 74.4 74.55 74.25
V3 1 76.05 82.35 89.1 97.05
V3 2 92.1 97.35 105 111.6
V3 3 75.9 82.2 90.75 97.5

1. Define the statistical linear model and define all terms of the statistical linear model.

Yijk=μ+ ρj +αi+ δij+ βk + ( αβ ) ik +εijk


where: i=1,2….a k=1,2…..b j=1,2….r

Yijk = Observation from subplot (Phosphorus levels) given the kth level of B in the main
plot (Variety) given the ith level of A in the jth block
μ= overall grand mean
ρj = effect of the jth block
αi = effect of the ith level of factor A (variety)
δij = random error associated with the main plot (variety) given the ith level of A in the jth
block
βk = effect of the kth level of factor B (phosphorus levels)
(αβ)ik = interaction effect between the ith level of A and the kth level of B
Εijk = random error associated with the subplot (phosphorus levels) given the kth level of
B in the main plot (variety) given the ith level of A in the Jth block

2. Propose a possible layout of the experiment. Justify the use of split plot design in this
experiment.

Justification on why split plot design


- It is suited for two-factor experiments which in this case is the variety and the
phosphorus.
- Higher precision for estimates of the effects of the subplot can be obtained using split
plot which corresponds to the interest, effect of the phosphorus levels, of the
researchers.
Proposed Layout

3. State the three hypotheses of interest (in terms of the parameters specified in your model).

Test for significance of interaction


Ho = There is no interaction between the variety and the levels of phosphorus ((αβ)ik = 0)
Ha = There is an interaction between the variety and the levels of phosphorus ((αβ)ik≠0)

Test for significance for variety


Ho = The mean yield of variety is the same for all levels of phosphorus (αi = o)
Ha = At least one variety has different mean yield (αi ≠ 0)

Test for significance of phosphorus levels


Ho = The mean yield is the same for all levels of phosphorus (βk = o)
Ha = at least one phosphorus level has a different mean yield (βk ≠ 0)

4. Carry out the analysis of variance and summarize your results in the ANOVA table.
Conduct the appropriate tests of hypotheses proposed in #3 and interpret your results.
Construct graphs to emphasize Define the scope of your conclusions based on whether
your model is fixed or random.
Conclusion for significance of interaction: Since P value (0.0000) < alpha significance of 5% (0.05),
we have enough evidence to say that there is an interaction between the variety and the levels of
phosphorus.

Conclusion for the significance of main plot: Since P value (0.0001) < alpha significance of 5%
(0.05), we have enough evidence to say that at least one variety has different mean yield.

Conclusion for the significance of subplot: Since P value (0.0000) < alpha significance of 5%
(0.05), we have enough evidence to say that at least one level of phosphorus has different mean
yield.

V2 exhibited the lowest yield, reaching its optimal level at P60. This observation is consistent with
the behavior of V1, which also displayed its highest yield at P60. However, both V2 and V1
experienced a decline in yield when the phosphorus levels exceeded P60. In contrast, V3's yield
demonstrated a continuous increase with higher phosphorus levels.

B. Twenty plots, each 10 ×4 meters, were randomly chosen in a large field of corn. For each
plot, the plant density (number of plants in the plot) and the mean cob weight (gm of
grain per cob) were observed. The results are given below.

Plant Density Cob Weight Plant Density Cob Weight


137 212 173 194
107 241 124 241
132 215 157 196
135 225 184 193
115 250 112 224
103 241 80 257
102 237 165 200
65 282 160 190
149 206 157 208
85 246 119 224

1. Construct a scatterplot with plant density at the x-axis and cob weight at the y-axis.
Describe the behaviour of the two variables relative to each other.

It can be observed that as plant density increases, the total cob weight decreases.

2. Calculate an appropriate measure of association to quantify the association. Test the


hypothesis that there is no linear association between plant density and cob weight.

Ho = There is no linear association between plant density and cob weight


Ha = there is a linear association between plant density and cob weight

r
tc=

−0.9418
√ 1−r 2
n−2
tc= =−11.88587608


2
1−(−0.9418)
20−2

Conclusion: Since tc > t value (2.101), cob weight is linearly associated with plant density

3. If you are to fit a regression line to describe the linear relationship between the two
variables. Cite the underlying basis of your model. Justify your choice of independent and
dependent variables.

- Since cob weight is linearly associated with plant density, it is the dependent variable
and plant density would be the independent variable. This implies that as the plant
density increases or decreases, the cob weight is expected to change in a consistent
and proportional manner. This suggests that higher plant densities may lead to larger
cob weights, while lower plant densities may result in smaller cob weights.

4. Fit a linear regression model based on your answer in #3. Do the computations manually
and write all solutions.

a) Calculate the y-intercept or the regression constant. Show your solution. Interpret the
value obtained in the context of the problem.

Xi = 137 + 107 + 132 + 135…+119 = 2561


Yi = 212 + 241 + 215 +225…+224 = 4482
XiYi = (137*212)+(107*241)…+(119*224) = 559357
X2 = 132^2 + 107^2…+119^2 = 348145
Y2 = 212^2 + 241^2…+224^2 = 1016248
SPXY = 559357-((2561*4482)/20) = -14563.1
SSX = 348145-(2561^2)/20) = 20208.95
SSY = 1016248-(4482^2)/20) = 11831.8
Sxy = -14563.1/20-1 = -766.48
Sx = square root of 20208.95/20-1 = 32.61
Sy = square root of 11831.8/20-1 = 24.95
r = -766.48 / (32.61*24.95) = -0.9418
β = -14563.1 / 20208.95 = -0.72063
α = 224.1 – (-0.72063*128.05) = 316.3762
S^2yx = 11831.8 – (-0.72063)(-14563.1)/20-2 = 74.3128

b) Test the hypothesis about the regression coefficient:

i) State the hypothesis about the y-intercept or the regression constant.


Ho: The regression constant is equal to zero
Ha: The regression constant is not equal to zero
ii) Test Statistic:

316.3762
tc= =39.5438


2
1 128.05
74.3128 +
20 20208.95

iv) Decision Rule: Since tc > t value, reject Ho

v) Conclusion and Interpretation:

Since the calculated t value is greater than the t tabulated value, we have enough
evidence to say that the regression constant is not equal to zero. This means if plant
density is lower, the cob weight will increase and vice versa.

5. Do you think your regression line will have good predicting ability? Justify.

2 (−0.7206262572)(−14563.1)
R= =0.8870
11831.8

The regression model exhibits a strong predictive capability, as indicated by the high R-
squared value of 0.8870. This value suggests that approximately 89% of the variability
observed in the data can be accounted for by the regression line.

ANOVA Regression Model

SV Df SS MS Fc
Regression 1 (-0.7206*- 10494.55225 (10494.55225/74.29154189)
14563.1) = = 141.2617369
10494.55225
Error 18 (11831.8 – (1337.247754/18)
10494.55225) = 74.29154189
=
1337.247754
Total 19 11831.8

Test of Hypothesis on β
Ho = cob weight is not linearly dependent on plant density
Ha = cob weight is linearly dependent on plant density
Conclusion: since Fc > F tabulated value (5.98), we have enough evidence to say that cob
weight is linearly dependent on plant density.

You might also like