Exercise 4

Differences in color and functional properties of isolated starches have been reported to be affected
by varietal differences. A researcher sought to characterize sorghum starches for potential
industrial applications. The data below are some functional and color properties of the isolated
starches measured by the research.
Variety Water Oil Swelling Solubility Colour
Absorption Absorption Power (%) L a b
Capacity Capacity (g/g)
(%) (%)
Dorado 159.86 127.98 9.93 5.99 86.15 0.65 7.35
Dorado 179.41 137.06 9.81 3.43 89.23 0.37 6.07
ICRISAT 139.26 124.79 10.13 2.00 80.95 3.85 4.22
ICRISAT 158.68 144.32 9.81 2.11 81.70 3.29 3.79
Kadaaga 138.89 88.65 10.36 5.83 75.75 3.56 3.50
Kadaaga 158.61 160.27 10.49 2.71 76.47 4.01 3.84
Kapaala 139.02 123.93 9.29 4.59 90.22 0.66 5.83
Kapaala 179.24 127.98 9.85 6.16 89.24 0.54 6.50

(a) Construct a summary statistics table for the dataset, indicating mean and standard error
values
(b) Is there any evidence that the functional properties of isolated starches is influenced by
differences in variety?
(c) The “L” value represents the Lightness of the starch samples where a value of 100
indicates the colour white and 0 indicated the colour black. Is there any evidence that the
lightness of the starch samples differs among varieties?
(d) Assign superscript letters to Table (a), indicating significant differences in mean values at
p<0.05. Which test did you use and why?

1
Solution
Table 1: Test for equal means for water absorption capacity (%) of different sorghum
starches
Sum of sqrs df Mean square F p (same)
Between groups: 592.299 3 197.433 0.5711 0.6633
Within groups: 1382.93 4 345.733 Permutation p
(n=99999)
Total: 1975.23 7 0.5364
Components of variance (only for random effects):
Var(group): -74.15 Var(error): 345.733 ICC: -0.273029
omega2: 0
Levene´s test for homogeneity of variance, from means p (same): 4.777E-59
Levene´s test, from medians p (same): 4.777E-59
Welch F test in the case of unequal variances: F=0.6279, df=2.182, p=0.658

Table 2: Test for equal means for oil absorption capacity (%) of different sorghum starches
Sum of sqrs df Mean square F p (same)
Between groups: 145.154 3 48.3847 0.069 0.9735
Within groups: 2804.85 4 701.212 Permutation p (n=99999)
Total: 2950 7 0.9533
Components of variance (only for random effects):
Var(group): -326.414 Var(error): 701.212 ICC: -0.870905
omega2: 0
Levene´s test for homogeneity of variance, from means p (same): 8.234E-62
Levene´s test, from medians p (same): 8.234E-62
Welch F test in the case of unequal variances: F=0.4519, df = 1.871, p=0.745

2
Table 3: Test for equal means for swelling power (g/g) of different sorghum starches
Sum of sqrs df Mean square F p (same)
Between groups: 0.753038 3 0.251013 4.489 0.09046
Within groups: 0.22365 4 0.0559125 Permutation p (n=99999)
Total: 0.976688 7 0.07509
Components of variance (only for random effects):
Var(group): 0.09755 Var(error): 0.0559125 ICC: 0.63566
omega2: 0.5668
Levene´s test for homogeneity of variance, from means p (same): 1.024E-56
Levene´s test, from medians p (same): 1.024E-56
Welch F test in the case of unequal variances: F=8.9, df=2.066, p=0.09786

Table 4: Test for equal means for solubility (%) different sorghum starches
Sum of sqrs df Mean square F p (same)
Between groups: 12.4173 3 4.13908 1.765 0.2926
Within groups: 9.3825 4 2.34563 Permutation p (n=99999)
Total: 21.7998 7 0.2735
Components of variance (only for random effects):
Var(group): 0.896729 Var(error): 2.34563 ICC: 0.276567
omega2: 0.2228
Levene´s test for homogeneity of variance, from means p (same): 9.625E-62
Levene´s test, from medians p (same): 9.625E-62
Welch F test in the case of unequal variances: F=4.458, df=1.675, p=0.2215

3
Table 5: Test for equal means for colour “L” different sorghum starches
Sum of sqrs df Mean square F p (same)
Between groups: 231.058 3 77.0193 53.45 0.001102
Within groups: 5.76385 4 1.44096 Permutation p (n=99999)
Total: 236.822 7 0.00774
Components of variance (only for random effects):
Var(group): 37.7892 Var(error): 1.44096 ICC: 0.963269
omega2: 0.9516
Levene´s test for homogeneity of variance, from means p (same): 2.029E-56
Levene´s test, from medians p (same): 2.029E-56
Welch F test in the case of unequal variances: F=106.9, df=2.138, p=0.007148

a) Construct a summary statistics table for the dataset, indicating mean and standard error values
Answer
Table 6: Summary Statistics for functional properties of isolated starches of different
varieties of sorghuma
Variety of Water Absorption Oil Absorption Swelling Solubility Colour
Sorghum Capacity (%) Capacity (%) Power (g/g) (%) L
Dorado 169.635 ± 9.775b 132.55 ± 4.54b 9.87 ± 0.06b 4.71 ± 1.28b 87.69 ±1.54b
ICRISAT 148.97 ± 9.71b 134.555 ± 9.765b 9.97 ± 0.16b 2.055 ± 0.055b 81.325 ± 0.375c
Kadaaga 148.75 ± 9.86b 124.46 ± 35.81b 10.425 ±0.065b 4.27 ± 1.56b 76.11 ± 0.36d
Kapaala 159.13 ± 20.11b 125.955 ± 2.025b 9.57 ± 0.28b 5.375 ± 0.785b 89.73 ± 0.49bd
a
Values are mean ± standard error for duplicate determination
b
p ˃0.05, no significant difference between functional properties of isolated starches of different
sorghum varieties
cd
p<0.05, there is significant difference between the functional properties of isolated starches of
different sorghum varieties

4
b) Is there any evidence that the functional properties of isolated starches is influenced by
differences in variety?
Answer
From the dataset given, there is no evidence that shows that the functional properties of isolated
starches is influenced by differences in variety of sorghum because the p - values obtained for the
analysis of variance were greater than the confidence level set at 95% (p<0.05).

c) The “L” value represents the Lightness of the starch samples where a value of 100 indicates the
colour white and 0 indicated the colour black. Is there any evidence that the lightness of the starch
samples differs among varieties?
Answer
The mean values obtained as shown in table 6, gives the evidence that the lightness of the starch
samples of sorghum differs among varieties since the values obtained were different for all the
varieties of sorghum.

d) Assign superscript letters to Table (a), indicating significant differences in mean values at
p<0.05. Which test did you use and why?
Answer
The test used was analysis of variance (ANOVA), a statistical tool used to test the equality of three
or more population means by analyzing samples variance. Tukey pairwise which is a component
of ANOVA helps compare the actual difference or equality that exist among samples.