You are on page 1of 2

PEER-R EV IEW ED

One-Way Analysis of Variance


Abhishek Vijayvargiya

INTRODUCTION For the purpose of this paper, we consider the following


Analysis of variance (ANOVA) is an extremely important sample problem (3).
method in exploratory and confirmatory data analysis.
Unfortunately, in some complex problems it is not always A SAMPLE PROBLEM
easy to set up an appropriate ANOVA. ANOVA is classified A medical device manufacturing company runs three
as one-way ANOVA and two-way ANOVA (1). This paper injection-molding machines for the production of silicone
shows how easily one-way ANOVA can be used to deter- valves. There is reason to believe that the three molding
mine whether there is a statistically significant difference processes may not be producing similar products. A
in the data analysis of a problem. company expert in designed experiments and statistical
techniques is called in to assess the operation. The expert
ANOVA—TERM REVEALED is presented with the data in Table I, which represent
In statistics, ANOVA is a collection of statistical models, the “Shore A” hardness, or durometer, of the molded
and their associated procedures, in which the observed silicone valves.
variance is partitioned into components because of dif- The problem can be easily solved using one-way
ferent explanatory variables. The initial techniques of the ANOVA and, after getting the final result, one can
ANOVA were developed by the statistician and geneticist compare it with the value obtained from an F Table
R. A. Fisher; thus, one can say that ANOVA is a statistical (2) for an alpha value of 0.05.
model meant to analyze data. Usually the variables in
an ANOVA analysis are categorical, not continuous. For FORMULAS USED
one-way ANOVA we make use of an F Table (2) with the For solving one-way ANOVA, one needs to find out the
value of alpha as 0.05. mean of the total sample. This can be done by adding the
values of all the components and dividing the total sum
ANOVA’S ROLE by total number of components. Further, one needs to
ANOVA for balanced data does the following three things find out the correction factor, which can be obtained by
at once: using Formula 1.
• Comparisons of mean squares, along with F-tests (1,
2)
• Under ANOVA the sum of squares indicate the variance [Formula 1]
of each component of the decomposition (1)
• Closely related to the ANOVA is a linear model fit with For the problem presented in this paper, after finding
coefficient estimates and standard errors. the correction factor we find the sum of squares (SOS)
or machine sum of squares (MSOS), under which we
IS ANOVA OBSOLETE? individually take the squares of the sum of all the three
What is the analysis of variance? Econometricians see it as machines and then add all the three outputs, and fur-
an uninteresting special case of linear regression. Instruc- ther subtract it from the correction factor obtained (1).
tors see it as one of the hardest topics in classical statis- The total sum of squares (TSOS) is found by using For-
tics to teach. However this paper shows how the ideas mula 2. Error can be then found by subtracting MSOS
of ANOVA are useful in many applications of statistics. by TSOS.

For more Author


information,
go to
gxpandjvt.com/bios
[ ABOUT THE AUTHOR
Abhishek Vijayvargiya is a graduate student in the Mechanical Engineering Department at the
University of Miami. He can be reached by e-mail at abhi1617@yahoo.co.in.

2 JOURNAL OF VALIDATION TECHNOLOGY [WINTER 2009] iv thome.com


A B H I S H E K V I J A Y VA R G I Y A

Table I: “Shore A” hardness, or durometer, of the


molded silicone valves.
Machine A Machine B Machine C
[Formula 2] 40 39.8 40.5

40.5 39.8 41
PROBLEM SOLVING
Table II presents the ANOVA for solving the sample 40 40.5 41
problem.
40.6 40 41.1
Machine degrees of freedom (DOF) = Number of
Machines – 1 40.2 41 41.2
Error DOF = Total number of data values - Number
of Machines
Total DOF = Total number of data values - 1 Table II: ANOVA for sample problem.
Machine A Machine B Machine C
Correction factor = 24579.46
Machine sum of squares = 1.732 40 39.8 40.5
Total sum of squares = 3.424 40.5 39.8 41
Error SOS- (3.424 – 1.732) = 1.692
40 40.5 41
Machine mean square = (Machine SOS)/(Machine
DOF) 40.6 40 41.1
Error mean square = (Error SOS)/(Error DOF)
40.2 41 41.2
Calculated F value = (Machine mean square)/(Error
mean square). Sum 201.3 201.1 204.8

Mean 40.26 40.22 40.96


Use F Table with alpha as 0.05 and degrees of
freedom as 12 and 2. The Tabled F value can also be
obtained from Excel using the following function: Table III: Final ANOVA.
SOS DOF MEAN Square F Value
Tabled F value = FINV(0.05, Machine DOF, Error
DOF). Machine 1.732 2 0.866 6.1418

Error 1.692 12 0.141


Calculate the F value as above.
TOTAL 3.424 14
If the Tabled F value is more than the calculated value
we fail to reject the hypothesis of equality of the means.
If the table value is less than the calculated value we reject • If the calculated F value is not less than the F tabular
the hypothesis of equality of the means. Table III is the value, reject the hypothesis of equality of the means.
final ANOVA table.
Because the calculated F value is 6.1418 and the value REFERENCES
obtained from F Table with 2 and 12 degrees of freedom 1. Hicks, Charles Robert and Turner, Kenneth V., Jr., Fundamen-
is 3.8853, which is less, we reject the hypothesis. tal Concepts in the Design of Experiments, New York: Oxford,
58-64, 507-520, 1999.
CONCLUSIONS 2. Table of F-Statistics P=0.05, http://www.statsoft.com/text-
The problem presented in this paper clearly shows how book/sttable.html.
ANOVA can be helpful in solving various statistical prob- 3. Gelman, Andrew, Analysis of Variance—Why It Is More Impor-
lems. The following should be kept in mind while using tant Than Ever, Columbia University, 2005. JVT
ANOVA:
• If you obtain a negative value when calculating quanti- ARTICLE ACRONYM LISTING
ties in ANOVA, which should be positive such as the ANOVA Analysis of Variance
SOS, Mean square or F, check your work DOF Degrees of Freedom
• Be sure to read the recommended references in this MSOS Machine Sum of Squares
paper before attempting an ANOVA to be sure your SOS Sum of Squares
data meet the assumptions (normality, variance homo-
geneity, independence, balance) of ANOVA
JOURNAL OF VALIDATION TECHNOLOGY [WINTER 2009] 3

You might also like