Slide 1

Chaughule, Thorat - Statistical Analysis/Design of Experiments

Figure 8.4. Effect of hot air drying on color values of Sapota pulp (Data from Chaughule
and Thorat, Formulation, drying & nutritional evaluation of ready to eat sapota {Achras
zapota} extrudes), submitted to IDS2010, proceedings)
Subsequent to a conclusion such as this, researchers often wish to determine which
of several pulp properties (e.g., protein content, total sugars, etc) contribute to the re-
duction and slight increase in L, a, b values. The difficulty with such a determination is
that the pulp properties of interest cannot all be specifically controlled in the dehydra-
tion of Sapota pulp. Because of this, many pulp properties that might subsequently be of
interest simultaneously vary across the test runs, resulting in a confounding of their ef-
fects.

This experiment was specifically designed only to investigate the effects on L, a, b
values during processing. Studies could be specifically designed to study some of the
other product properties. In such studies the pulp properties would be varied in a sys-
tematic fashion and confounding among the properties could be eliminated by the
choice of the design. Note that in this example it is the uncontrolled variation of the pulp
properties that leads to the confounding of their effects on the response.

8.7.3. Incorrect principles of efficiency lead to unnecessary or inconclusive

results

The aforementioned sapota pulp example demonstrates the need to construct de-
signs in which factors of interest are systematically varied and need to consider the like-
ly magnitude of the inherent variation in the test results when planning the number of
test runs. In academic research sometime the desire to run economical experiments can
lead to strategies that may in fact be wasteful or even fail to achieve the expected goals
of the project. Time and cost efficiencies are always important objectives in experimen-
tal work. Occasionally efficiency becomes an overriding consideration and the project
goals become secondary. If time or budgetary considerations lead to undue restrictions
on the factors and levels that can be investigated, the project goals should be reeva-

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0

5

10

15

20

25

30

35

40

45

0 50 100 150 200 250 300

Rate of drying kg kg

-^1 min
-^1

Colour Value

Time (min)

L a b Drying curve