Strategy of Experimentation

Experiment is a Test.
 Purposeful changes are made to input
variables to identify and observe the reasons
for changes in output response.
Objectives
I. New product design
II. Manufacturing process development
III. Process improvement
 Strategy of Experimentation
Examples

Metallurgical Engineer
I. Oil quenching vs Salt quenching (to produce maximum
hardness for alloy)

Statistician
I. Deterministic vs Probabilistic (to capture behavior of variables)

 Mechanical Engineer
I. Running turbine at various speed to find maximum efficiency
Metallurgical Engineer example

Importance of Data collection Procedure

Results and conclusions depend on the manner in which data were collected
Don’t make data.
Strategy of experimentation
 The general approach to planning and conducting the experiment is called
the strategy of experimentation.

Example: A golfer wants to lower his score. Variables that may influence his
golf score are as follows:

1. The type of driver used (oversized or regular sized)

2. The type of ball used (balata or three piece)
3. Walking and carrying the golf clubs or riding in a golf cart
4. Drinking water or drinking “something else” while playing
5. Playing in the morning or playing in the afternoon
6. Playing when it is cool or playing when it is hot
7. The type of golf shoe spike worn (metal or soft)
8. Playing on a windy day or playing on a calm day.
 Strategy of experimentation
One-factor-at-a-time approach
 The OFAT method consists of selecting a starting point, or
baseline set of levels, for each factor, and then successively
varying each factor over its range with the other factors held
constant at the baseline level. After all tests are performed, a
series of graphs are usually constructed showing how the
response variable is affected by varying each factor with all
other factors held constant.

O= Oversized B= Balata W= walk W= Water

R= Regular-sized T= Three-piece R= Riding SE= something else
 Strategy of experimentation
Conclusions
Using these one-factor-at-a-time graphs, we would select the optimal
combination to be the regular-sized driver, riding, and drinking water.
The type of golf ball seems unimportant.
Disadvantages
The major disadvantage of the OFAT strategy is that it fails to consider any
possible interaction between the factors.
 An interaction is the failure of one factor to produce the same effect on
the response at different levels of another factor.

Factorial experiment
•This is an experimental strategy in which factors are varied together,
instead of one at a time.
A two-factor factorial experiment
A two-factor factorial experiment
Conclusion:
•That is, on average, switching from the oversized to the
regular-sized driver increases the score by 3.25 strokes per
round.

•The results of this factorial experiment indicate that driver

effect is larger than either the ball effect or the interaction.

•Statistical testing could be used to determine whether any of

these effects differ from zero. In fact, it turns out that there is
reasonably strong statistical evidence that the driver effect
differs from zero and the other two effects do not.

•Therefore, this experiment indicates that Golfer should

always play with the oversized driver.
A three-factor factorial experiment

 3 factors and at level 2
• There are eight test combinations of these three factors
across the two levels of each.
•These eight trials can be represented geometrically as
the corners of a cube.
A four-factor factorial experiment

•Generally, 16 runs required.
•A fractional factorial experiment is a variation of
the basic factorial design in which only a subset of
the runs is used.
Basic Principles of experimental design
Statistical Design of experiments:
The process of planning the experiment so that appropriate
data will be collected and analyzed by statistical methods,
resulting in valid and objective conclusions.

There are two aspects to any experimental problem.

1. The design of the experiment
2. The statistical analysis of the data.

The three basic principles of experimental design are:

1. Randomization
2. Replication
3. Blocking
Basic Principles of experimental design
 Randomization:
•Randomization is both the allocation of the experimental material and the
order in which the individual runs of the experiment are to be performed are
randomly determined.

•By properly randomizing the experiment, we assist in “averaging out” the

effects of extraneous factors.

•For example, suppose that the specimens are of slightly different

thicknesses. If all the specimens subjected to the oil quench are thicker than
those subjected to the saltwater quench, we may be introducing systematic
bias into the experimental result.

•Computer software programs are widely used to assist experimenters in

selecting and constructing experimental designs. [Uniformity and
Independence].
Basic Principles of experimental design
 Replication:
•Replication is an independent repeat run of each factor combination.
•In the metallurgical, treating a specimen by oil quenching or by saltwater
quenching is a replication.
•If five specimens are treated in each quenching medium, five replicates have
been obtained.
Each of the 10 observations should be run in random order.
Replication allows the experimenter to obtain an estimate of the
experimental error (i.e., human error, data entry mistakes, etc).
Experimental error is the difference between a measurement and the true
value or between two measured values.
Replication permits the experimenter to obtain a more precise estimate of
this parameter (i.e., mean). As one experiment is not enough to compare
means in the metallurgical example.
 Basic Principles of experimental design
 Blocking:
•Blocking is a design technique used to improve the precision with
which comparisons among the factors of interest are made.
•Often blocking is used to reduce or eliminate the variability
transmitted from nuisance factors—that is, factors that may influence
the experimental response but in which we are not directly interested.
Example
• An experiment in a chemical process may require two batches of raw
material to make all the required runs. There could be differences
between the batches due to supplier-to-supplier variability.
•Generally, a block is a set of relatively homogeneous experimental
conditions. Hence, in the chemical process example, each batch of raw
material would form a block.
 Basic Principles of experimental design
Accuracy: Experimental error is the difference between a
measurement and the true value or between two measured values.
Precision: Accuracy measures how close a measured value is to the
true value or accepted value.
Using Statistical Techniques in Experimentation
1. Use your nonstatistical knowledge of the problem.
• utilizing expert judgment elicitation.

2. Keep the design and analysis as simple as possible.

• Simple the better.
• Elegant statistics can not replace a bad design.

3. Recognize the difference between practical and statistical

significance.
• If new design improves efficiency of a turbine (say, by only 1% )and cost
is half of the price. 1% improvement is probably too small to be any
practical value.

4. Experiments are usually iterative.

• Don’t design too comprehensive an experiment at the start of study. As
we are not well-equipped to answer complex questions in the beginning.
• Experiments require repetitions. Don’t spend more than 25% of budget
in the initial experiment.