You are on page 1of 60

Research Method

Topic: The effects of type of starches


and plasticizers on tensile properties
Group members:
Nguyen Vu Hoang Phuong
Suttikan Saipiroontong
Papassara Bumroongsak
Puntipa Phongteeramit
Introduction
Edible film
Renew resources
Polysaccharide
Protein
Lipid
Biodegradable
Inexpensive
Enhance and protect food product
Environmental pollution
Hypothesis
Cassava starch could be used to replace Mungbean starch as mush as 50%.
Both type of starches and plasticizer will have an effect on mechanical properties.

Objective
To study the AM content of different type of starches
To study effects of type of starches (AM content) and plasticizers on mechanical,
properties of starch films.
Result and Discussion
Factor =Type of starch
(Cassava, Mung Bean and Cassava +Mung Bean)
Factor level (# treatment) (a) = 3 level
Factor =Type of form
(powder and Film)
Factor level (# treatment) (b) = 2 level
Design : Factorial in CRD
Replication (n) = 2 replicates
Response = AP IM and AM
Number of observation/runs = 322=12
Hypothesis Testing

H0 : 1=2==a=0 VS H1: at least one I 0


H0 : 1=2== b =0 VS H1: at least one j0
H0 : ()ab =0 VS H1 :at least one ()ij 0

Where
I = the effect of type of starch i (i = 1,2,3)
j = the effect of form of starch j (j=1,2)
()ij = the effect of interaction between I and
j
Analysis of Variance of AP

Source DF Adj SS Adj MS F-Value P-Value


starch 2 225.97 112.98 7.39 0.024
form 1 195.99 195.99 12.82 0.012
starch*form 2 232.57 116.28 7.61 0.023
Error 6 91.74 15.29
Total 11 746.26

Since p-value of starch form and interaction between


starch*form is less than 0.05 , were reject the null hypothesis
at =0.05. It shows all factor has effect on AP.
The Normality Assumption

H0 : the error is normally distributed


H1 : the error is not normally
distributed

P-value is greater than 0.05,


were accept the null hypothesis. It
means that the normality
assumption is valid.
The outlier

There is no outlier in the data.

Largest residual is e32 d32 = 3.930/ 15.29=1.005


|d32| < 3
The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Comparisons for AP
Fisher Pairwise Comparisons: Response = AP, Term = Starch

Grouping Information Using Fisher LSD Method and 95%


Confidence

Starch N Mean Grouping


2 4 47.7387 A
3 4 46.6742 A
1 4 38.0475 B

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs

Fisher Pairwise Comparisons: Response = AP, Term = Form

Grouping Information Using Fisher LSD Method and 95%


Confidence

Form N Mean Grouping


2 6 48.1948 A
1 6 40.1122 B

Means that do not share a letter are significantly different.


Comparisons for AP
Fisher Pairwise Comparisons: Response = AP, Term = Starch*Form

Grouping Information Using Fisher LSD Method and 95%


Confidence

Starch N Mean Grouping


2 4 47.7387 A
3 4 46.6742 A
1 4 38.0475 B

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs
Fisher Pairwise Comparisons: Response = AP, Term = starch

Grouping Information Using Fisher LSD Method and 95%


Confidence

Starch*Form N Mean Grouping


2 2 2 50.7140 A
1 2 2 47.9340 A
3 1 2 47.4120 A
3 2 2 45.9365 A
2 1 2 44.7635 A
1 1 2 28.1610 B

Means that do not share a letter are significantly different.


Analysis of Variance of IM

Source DF Adj SS Adj MS F-Value P-Value

Starch 2 1295.16 647.580 165.47 0.000


Form 1 230.41 230.414 58.87 0.000
Starch*Form 2 255.89 127.946 32.69 0.001
Error 6 23.48 3.914
Total 11 1804.95

Since p-value of starch form and interaction between


starch*form is less than 0.05 , were reject the null
hypothesis at =0.05. It shows all factor has effect on AP.
The Normality Assumption

H0 : the error is normally


distributed
H1 : the error is not normally
distributed

P-value is greater than 0.05,


were accept the null hypothesis. It
means that the normality assumption
is valid.
The outlier

There is no outlier in the data.

Largest residual is e32 d32 = 1.782/ 3.914=0.901


|d32| < 3
The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant
variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Comparisons for IM
Fisher Pairwise Comparisons: Response = IM, Term = Starch

Grouping Information Using Fisher LSD Method and 95%


Confidence

Starch N Mean Grouping


1 4 37.7505 A
3 4 18.1582 B
2 4 3.8905 C

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs
Fisher Pairwise Comparisons: Response = IM, Term = Form

Grouping Information Using Fisher LSD Method and 95%


Confidence

Form N Mean Grouping


1 6 27.6483 A
2 6 18.8845 B

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs
Fisher Pairwise Comparisons: Response = IM, Term = Form

Grouping Information Using Fisher LSD Method and 95%


Confidence

Form N Mean Grouping


1 6 27.6483 A
2 6 18.8845 B

Means that do not share a letter are significantly different.


Analysis of Variance of AM

Source DF Adj SS Adj MS F-Value P-Value


starch 2 441.665 220.832 11.25 0.009
form 1 1.391 1.391 0.07 0.799
starch*form 2 0.584 0.292 0.01 0.985
Error 6 117.797 19.633
Total 11 561.436

Since p-value of starch is less than 0.05 , were reject the null
hypothesis at =0.05. It shows all factor has effect on AM
The Normality Assumption

H0 : the error is normally


distributed
H1 : the error is not normally
distributed

P-value is greater than 0.05,


were accept the null hypothesis. It
means that the normality assumption
is valid.
The outlier

There is no outlier in the data.

Largest residual is e12 d12 = 5.145/ . =1.161


|d12| < 3
The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Comparisons for AM
Fisher Pairwise Comparisons: Response = AM, Term = Starch

Grouping Information Using Fisher LSD Method and 95%


Confidence

Starch N Mean Grouping


2 4 38.3707 A
3 4 35.1670 A
1 4 24.2020 B

Means that do not share a letter are significantly different.


Interaction plot of type of starch and percent of component
Result and Discussion
Factor =Type of starch
(Cassava, Mung bean and Cassava + Mung bean)
Factor level (# treatment) (a) = 3 level
Factor =Type of plasticizer
(Glycerol and Sorbitol)
Factor level (# treatment) (b) = 2 level
Design : Factorial in CRD
Replication (n) = 2 replicates
Response = Stress , Modulus , Thickness and Strain
Number of observation/runs = 322=12
Hypothesis Testing

H0 : 1=2==a=0 VS H1: at least one I 0


H0 : 1=2== b =0 VS H1: at least one j0
H0 : ()ab =0 VS H1 :at least one ()ij 0

Where
I = the effect of type of starch i (i = 1,2,3)
j = the effect of type of plasticizer j (j=1,2)
()ij = the effect of interaction between I and
j
Analysis of Variance of Stress

Source DF Adj SS Adj MS F-Value P-Value


starch 2 210.509 105.254 283.26 0.000
plasticizer 1 267.474 267.474 719.83 0.000
starch*plasticizer 2 24.547 12.274 33.03 0.001
Error 6 2.229 0.372
Total 11 504.759

Since p-value of starch plasticizer and interaction between


starch*plasticizer is less than 0.05 , were reject the null
hypothesis at = 0.05, It shows all factor has effect on stress.
The Normality Assumption

H0 : the error is normally distributed


H1 : the error is not normally distributed

P-value is greater than 0.05,


were accept the null hypothesis It
means that the normality
assumption is valid.
The Outlier

There is no outlier in the data.

Largest residual is e21


d21= 0.978/ 0.372=1.606
|d21| < 3
The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Fisher Individual 95% CIs
Fisher Pairwise Comparisons: Response = Stress, Term = Starch, plasticizer

Grouping Information Using Fisher LSD Method and 95% Confidence

Starch N Mean Grouping


2 4 14.9887 A
3 4 13.7448 B
1 4 5.5474 C

Plasticizer N Mean Grouping


2 6 16.1482 A
1 6 6.7058 B

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs
Fisher Pairwise Comparisons: Response = Stress, Term = Starch*Plasticizer

Grouping Information Using Fisher LSD Method and 95% Confidence

starch*plasticizer N Mean Grouping


2 2 2 20.6369 A
3 2 2 19.5594 A
2 1 2 9.3406 B
1 2 2 8.2482 B
3 1 2 7.9302 B
1 1 2 2.8467 C

Means that do not share a letter are significantly different.


Analysis of Variance of Modulus

Source DF Adj SS Adj MS F-Value P-Value


starch 2 1524537 762269 7.37 0.024
plasticizer 1 1335929 1335929 12.91 0.011
starch*plasticizer 2 79711 39856 0.39 0.696
Error 6 620665 103444
Total 11 3560843

Since p-value of starch and plasticizer is less than 0.05 ,


were reject the null hypothesis at =0.05. It shows starch and
plasticizer has effect on modulus.
The Normality Assumption

P-value is greater than 0.05,


were accept the null hypothesis
It means that the normality
assumption is valid.

H0 : the error is normally distributed


H1 : the error is not normally
The outlier

There is no outlier in the data.

Largest residual is e31 d31 = 476.279/ 103444=1.481


|d31| < 3
The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Comparisons for Modulus
Fisher Pairwise Comparisons: Response = Modulus, Term = Starch

Grouping Information Using Fisher LSD Method and 95%


Confidence

Starch N Mean Grouping


3 4 1163.87 A
2 4 1044.27 A
1 4 355.09 B

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs

Fisher Pairwise Comparisons: Response = Modulus, Term = Plasticizer

Grouping Information Using Fisher LSD Method and 95%


Confidence

Plasticizer N Mean Grouping


2 6 1188.07 A
1 6 520.75 B

Means that do not share a letter are significantly different.


Analysis of Variance of Thickness

Source DF Adj SS Adj MS F-Value P-Value


starch 2 0.000058 0.000029 2.35 0.176
plasticizer 1 0.000231 0.000231 18.73 0.005
starch*plasticizer 2 0.000151 0.000075 6.12 0.036
Error 6 0.000074 0.000012
Total 11 0.000513

Since P-value of plasticizer and interaction between


starch*plasticizer is less than 0.05 , were reject the null
hypothesis at =0.05 , It shows plasticizer and interaction
between starch*plasticizer has effect on thickness.
The Normality Assumption

P-value is greater than 0.05 ,


were accept the null hypothesis It
means that the normality assumption
is valid.

H0 : the error is normally distributed


H1 : the error is not normally distributed
The outlier

There is no outlier in the data.

Largest residual is e32 d32 = 0.0039/ 0.000012=1.114


|d32| < 3
The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Comparisons for Thickness
Fisher Pairwise Comparisons: Response = Thickness, Term = Plasticizer

Grouping Information Using Fisher LSD Method and 95%


Confidence

Plasticizer N Mean Grouping


2 6 0.106544 A
1 6 0.097778 B

Means that do not share a letter are significantly different.


Fisher Individual 95% CIs
Fisher Pairwise Comparisons: Response = Thickness, Term = Starch*Plasticizer

Grouping Information Using Fisher LSD Method and 95%


Confidence

starch*plasticizer N Mean Grouping


2 2 2 0.111567 A
3 2 2 0.105800 A B
1 1 2 0.103500 A B
1 2 2 0.102267 B
2 1 2 0.097267 B C
3 1 2 0.092567 C

Means that do not share a letter are significantly different.


Analysis of Variance of Strain

Source DF Adj SS Adj MS F-Value P-Value


starch 1 2.211 2.211 2.92 0.162
plasticizer 1 179.943 179.943 237.89 0.000
starch*plasticizer 1 2.009 2.009 2.66 0.178
Error 4 3.026 0.756
Total 7 187.189

Since p-value of plasticizer is less than 0.05 , were reject the


null hypothesis at =0.05. It shows plasticizer has effect on strain.
The Normality Assumption

H0 : the error is normally distributed


H1 : the error is not normally
distributed

P-value is greater than 0.05 ,


were accept the null hypothesis. It
means that the normality assumption
is valid.
The outlier

Largest residual is e22 d22 = 0.845/ 0.756=0.972


|d22| < 3

There is no outlier in the data.


The Independent Assumption

Assumption is valid (does not show any pattern)


The Constant Variance Assumption

H0 : constant variance
H1 : non constant variance

Shows spread at one end than the other (megaphone)


So, the constant variance assumption is not valid.
Comparisons for Strain
Fisher Pairwise Comparisons: Response = Strain, Term = Plasticizer

Grouping Information Using Fisher LSD Method and 95%


Confidence

Plasticizer N Mean Grouping


1 4 21.3502 A
2 4 11.8649 B

Means that do not share a letter are significantly


different.
Interaction plot of starches and plasticizers on mechanical
properties
Conclusions:

The tensile properties of starch film strongly depend on type of starches and
plasticizers.
The AM content has significantly affected on mechanical
The AM content of the mixed starch and mungbean starch has no siginifcantly
different.

You might also like