You are on page 1of 8

Statistical Tool for Agricultural Research (STAR)

Mon Nov 25 21:12:35 2019

Analysis of Variance
Completely Randomized Design

====================================================
ANALYSIS FOR RESPONSE VARIABLE: MoistureContent
====================================================

Summary Information
------------------------------------------------
FACTOR NO. OF LEVELS LEVELS
------------------------------------------------
Treatment 5 L1, L2, L3, L4, L5
------------------------------------------------
Number of Observations Read and Used: 20

Descriptive Statistics
\---------------------------------------------------------
Variable N_Obs Min Max Mean StdDev
---------------------------------------------------------
MoistureContent 20 4.10 77.40 27.28 21.01
---------------------------------------------------------

Test for Homogeneity of Variances


-------------------------------------------
Method DF Chisq Value Pr(>Chisq)
-------------------------------------------
Bartlett 4 8.07 0.0891
-------------------------------------------

Test for Normality


----------------------------------------------------------
Variable Method W Value Pr(< W)
----------------------------------------------------------
MoistureContent_resid Shapiro-Wilk 0.8779 0.0162
----------------------------------------------------------

ANOVA TABLE
Response Variable: MoistureContent
------------------------------------------------------------
Source DF Sum of Square Mean Square F Value Pr(> F)
------------------------------------------------------------
Treatment 4 716.5120 179.1280 0.35 0.8398
Error 15 7672.7200 511.5147
Total 19 8389.2320
------------------------------------------------------------

Summary Statistics
------------------------------
CV(%) MoistureContent Mean
------------------------------
82.91 27.28
------------------------------
Standard Errors
--------------------
Effects StdErr
--------------------
Treatment 15.99
--------------------

Table of Means
-----------------------------------
Treatment MoistureContent Means
-----------------------------------
L1 30.30
L2 26.00
L3 16.30
L4 33.50
L5 30.30
-----------------------------------

INTERPRETATIONS

 FOR THE DESCRIPTIVE STATISTICS


The standard deviation of 21.01 implies that the distance of the treatment
means from the general mean of 27.28 is very far.

 FOR THE TEST FOR HOMOGENEITY OF VARIANCES


Method Used: Chisquare Test
The p-value of 0.0891 is greater than the alpha value of 0.05 (Pr >
Chisq) which implies that the variances are homogenous, hence it
accepts the null hypothesis.

 FOR THE TEST FOR NORMALITY


Method Used: Shapiro-Wilk Test
The p-values of 0.0162 is less than the alpha value of 0.05 (Pr < W) which
implies that the variances are not normally distributed, hence it rejects
the null hypothesis.

 FOR THE MULTIPLE COMPARISON TEST


Method Used: Least Significant Difference Test
The p-value of 0.8398 is greater than the alpha of 0.05, hence it accepts
the null hypothesis, indicating that there is no significant difference
in the means of all the treatments in terms of the moisture content of
the samples. Moreover, L4 obtained the mean of 33.50 which is the highest
moisture content mean of all the treatments. This implies that L4 is the
heaviest among all the samples because moisture content affects the mass
of the object in relation to the surface area. Despite the numerical
differences of the results, all the levels are not significantly different
with each other.
====================================================
ANALYSIS FOR RESPONSE VARIABLE: WaterAbsorption
====================================================

Summary Information
------------------------------------------------
FACTOR NO. OF LEVELS LEVELS
------------------------------------------------
Treatment 5 L1, L2, L3, L4, L5
------------------------------------------------
Number of Observations Read and Used: 20

Descriptive Statistics
-------------------------------------------------------
Variable N_Obs Min Max Mean StdDev
-------------------------------------------------------
WaterAbsorption 20 80 706 230.85 173.75
-------------------------------------------------------

Test for Homogeneity of Variances


-------------------------------------------
Method DF Chisq Value Pr(>Chisq)
-------------------------------------------
Bartlett 4 25.26 0.0000
-------------------------------------------

Test for Normality


----------------------------------------------------------
Variable Method W Value Pr(< W)
----------------------------------------------------------
WaterAbsorption_resid Shapiro-Wilk 0.8038 0.0010
----------------------------------------------------------

ANOVA TABLE
Response Variable: WaterAbsorption
------------------------------------------------------------
Source DF Sum of Square Mean Square F Value Pr(> F)
------------------------------------------------------------
Treatment 4 385284.8000 96321.2000 7.67 0.0014
Error 15 188311.7500 12554.1167
Total 19 573596.5500
------------------------------------------------------------

Summary Statistics
------------------------------
CV(%) WaterAbsorption Mean
------------------------------
48.54 230.85
------------------------------

Standard Errors
--------------------
Effects StdErr
--------------------
Treatment 79.23
--------------------
Pairwise Mean Comparison of Treatment

Least Significant Difference (LSD) Test

Alpha 0.05
Error Degrees of Freedom 15
Error Mean Square 12554.1167
Critical Value 2.1314
Test Statistics 168.8702

Summary of the Result:


----------------------------------
Treatment means N group
----------------------------------
L1 96.00 4 c
L2 330.00 4 ab
L3 182.75 4 bc
L4 449.50 4 a
L5 96.00 4 c
----------------------------------
Means with the same letter are not significantly different.

INTERPRETATIONS

 FOR THE DESCRIPTIVE STATISTICS


The standard deviation of 173.75 implies that the distance of the
treatment means from the general mean of 230.85 is very far.

 FOR THE TEST FOR HOMOGENEITY OF VARIANCES


Method Used: Chisquare Test
The p-value of 0.0000 is less than the alpha value of 0.05 (Pr <
Chisq) which implies that the variances are not homogenous, hence it
rejects the null hypothesis.

 FOR THE TEST FOR NORMALITY


Method Used: Shapiro-Wilk Test
The p-values of 0.0010 is less than the alpha value of 0.05 (Pr < W) which
implies that the variances are not normally distributed, hence it rejects
the null hypothesis.

 FOR THE MULTIPLE COMPARISON TEST


Method Used: Least Significant Difference Test
The p-value of 0.0014 is less than the alpha of 0.05, hence it rejects
the null hypothesis, indicating that there is a significant difference in
the means of all the treatments in terms of the water absorption of the
samples. Based on the ANOVA result, L4 gained the highest water absorption
mean given by 449.50, which is highly significant among all the
treatments. However, L4 is significantly closer to L2 but significantly
different with L1, L3, and L5. Moreover, L1 and L5 gained the same
significance albeit the numerical differences.
======================================================
ANALYSIS FOR RESPONSE VARIABLE: ThicknessSwelling
======================================================

Summary Information
------------------------------------------------
FACTOR NO. OF LEVELS LEVELS
------------------------------------------------
Treatment 5 L1, L2, L3, L4, L5
------------------------------------------------
Number of Observations Read and Used: 20

Descriptive Statistics
--------------------------------------------------------
Variable N_Obs Min Max Mean StdDev
--------------------------------------------------------
ThicknessSwelling 20 25 33 29.80 4.02
--------------------------------------------------------

Test for Homogeneity of Variances


-------------------------------------------
Method DF Chisq Value Pr(>Chisq)
-------------------------------------------
Bartlett 4 0.0000
-------------------------------------------

Test for Normality


------------------------------------------------------------
Variable Method W Value Pr(< W)
------------------------------------------------------------
ThicknessSwelling_resid Shapiro-Wilk 0.5212 0.0000
------------------------------------------------------------

ANOVA TABLE
Response Variable: ThicknessSwelling
------------------------------------------------------------
Source DF Sum of Square Mean Square F Value Pr(> F)
------------------------------------------------------------
Treatment 4 307.2000 76.8000 NaN NaN
Error 15 0.0000 0.0000
Total 19 307.2000
------------------------------------------------------------

INTERPRETATIONS

 FOR THE DESCRIPTIVE STATISTICS


The standard deviation of 4.02 implies that the distance of the treatment
means from the general mean of 29.80 is very far.

 FOR THE TEST FOR HOMOGENEITY OF VARIANCES


Method Used: Chisquare Test
The p-value of 0.0000 is less than the alpha value of 0.05 (Pr <
Chisq) which implies that the variances are not homogenous, hence it
rejects the null hypothesis.
 FOR THE TEST FOR NORMALITY
Method Used: Shapiro-Wilk Test
The p-values of 0.0000 is less than the alpha value of 0.05 (Pr < W) which
implies that the variances are not normally distributed, hence it rejects
the null hypothesis.

 FOR THE MULTIPLE COMPARISON TEST


Method Used: Least Significant Difference Test
Based on the result, the data of all the treatments cannot be subjected
to Analysis of Variance due to the characteristics of the gathered data
such as but not limited to: the data are not homogenous and not normally
distributed within the groups.
====================================================
ANALYSIS FOR RESPONSE VARIABLE: TensileStrength
====================================================

Summary Information
------------------------------------------------
FACTOR NO. OF LEVELS LEVELS
------------------------------------------------
Treatment 5 L1, L2, L3, L4, L5
------------------------------------------------
Number of Observations Read and Used: 20

Descriptive Statistics
--------------------------------------------------------
Variable N_Obs Min Max Mean StdDev
--------------------------------------------------------
TensileStrength 20 2.25 9.85 5.33 2.79
--------------------------------------------------------

Test for Homogeneity of Variances


-------------------------------------------
Method DF Chisq Value Pr(>Chisq)
-------------------------------------------
Bartlett 4 2.87 0.5796
-------------------------------------------

Test for Normality


----------------------------------------------------------
Variable Method W Value Pr(< W)
----------------------------------------------------------
TensileStrength_resid Shapiro-Wilk 0.9488 0.3493
----------------------------------------------------------

ANOVA TABLE
Response Variable: TensileStrength
------------------------------------------------------------
Source DF Sum of Square Mean Square F Value Pr(> F)
------------------------------------------------------------
Treatment 4 147.2870 36.8218 669.49 0.0000
Error 15 0.8250 0.0550
Total 19 148.1120
------------------------------------------------------------

Summary Statistics
------------------------------
CV(%) TensileStrength Mean
------------------------------
4.40 5.33
------------------------------

Standard Errors
--------------------
Effects StdErr
--------------------
Treatment 0.1658
--------------------
Pairwise Mean Comparison of Treatment

Least Significant Difference (LSD) Test

Alpha 0.05
Error Degrees of Freedom 15
Error Mean Square 0.0550
Critical Value 2.1314
Test Statistics 0.3535

Summary of the Result:


---------------------------------
Treatment means N group
---------------------------------
L1 2.52 4 d
L2 4.95 4 c
L3 7.10 4 b
L4 9.55 4 a
L5 2.52 4 d
---------------------------------
Means with the same letter are not significantly different.

INTERPRETATIONS

 FOR THE DESCRIPTIVE STATISTICS


The standard deviation of 2.79 implies that the distance of the treatment
means from the general mean of 5.33 is near.

 FOR THE TEST FOR HOMOGENEITY OF VARIANCES


Method Used: Chisquare Test
The p-value of 0.5796 is greater than the alpha value of 0.05 (Pr <
Chisq) which implies that the variances are homogenous, hence it
accepts the null hypothesis.

 FOR THE TEST FOR NORMALITY


Method Used: Shapiro-Wilk Test
The p-values of 0.3493 is greater than the alpha value of 0.05 (Pr > W)
which implies that the variances are normally distributed, hence it
accepts the null hypothesis.

 FOR THE MULTIPLE COMPARISON TEST


Method Used: Least Significant Difference Test
The p-value of 0.0000 is less than the alpha of 0.05, hence it rejects
the null hypothesis, indicating that there is a significant difference in
the means of all the treatments in terms of the tensile strength of the
samples. Based on the ANOVA result, L4 has the highest mean of tensile
strength among all the treatments which is 9.55. L4 is very significantly
different among all the treatments, yet L1 and L5 are not significantly
different with each other.

You might also like