See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/283120377

Modeling and Statistical Analysis of a Soap Production Mix in Bejoy

Manufacturing Industry, Anambra State, Nigeria

Article · January 2015

CITATIONS READS

0 3,602

4 authors, including:

Okolie Paul Chukwulozie Jude Sinebe

Nnamdi Azikiwe University Delta State University
82 PUBLICATIONS 253 CITATIONS 40 PUBLICATIONS 245 CITATIONS

SEE PROFILE SEE PROFILE

Enyi Chukwudi Louis

Delta State Polytechnic Ogwashi-uku
2 PUBLICATIONS 1 CITATION

SEE PROFILE

All content following this page was uploaded by Okolie Paul Chukwulozie on 24 October 2015.

The user has requested enhancement of the downloaded file.

International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 454
ISSN 2229-5518

Modeling and Statistical Analysis of a Soap

Production Mix in Bejoy Manufacturing Industry,
Anambra State, Nigeria
Okolie Chukwulozie Paul, Iwenofu Chinwe Onyedika,Sinebe Jude Ebieladoh and Enyi Chukwudi Louis

Abstract - The research work is based on the statistical analysis of the processing data. The essence is to analyze the data statistically
and to generate a design model for the production mix of soap manufacturing products in Bejoy manufacturing company Nkpologwu,
Aguata Local Government Area, Anambra state, Nigeria. The statistical analysis shows the statistical analysis and the correlation of the
data. T test, Partial correlation and bivariate correlation were used to understand what the data portrays. The design model developed was
used to model the data production yield and the correlation of the variables show that the R2 is 98.7%. However, the results confirm that
the data is fit for further analysis and modeling. This was proved by the correlation and the R-squared.

Index Term – General Linear Model, Correlation, Variables, Pearson, Significance, T-test, Soap, Production Mix and Statistic

——————————  ——————————
1. Introduction
Statistics is the study of the collection, organization, interpretation or explanation, and presentation of

IJSER
analysis, interpretation and presentation of data (Dodge data, (while others consider it a branch of
Y., 2006). It deals with all aspects of data, including the
mathematics (Moore David, 1992) concerned with
planning of data collection in terms of the design of
collecting and interpreting data. Because of its
surveys and experiments.
The word statistics, when referring to the scientific empirical roots and its focus on applications,

discipline, is singular, as in "Statistics is an art” (Moses, statistics is usually considered a distinct

1986). This should not be confused with the word mathematical science rather than a branch of
statistic, referring to a quantity (such as mean or
mathematics (Chance et al, 2005 & Anderson et al
median) calculated from a set of data, (Hays & William
1994). Much of statistics is non-mathematical:
Lee, 1973) whose plural is statistics ("this statistic
ensuring that data collection is undertaken in a
seems wrong" or "these statistics are misleading").
More probability density is found the closer one gets to way that produces valid conclusions; coding and

the expected (mean) value in a normal distribution. archiving data so that information is retained and
Statistics used in standardized testing assessment are made useful for international comparisons of
shown. The scales include standard deviations,
official statistics; reporting of results and
cumulative percentages, percentile equivalents, Z-
summarized data (tables and graphs) in ways
scores, T-scores, standard nines, and percentages in
comprehensible to those who must use them;
standard nines.
The objective of the study is to extract information implementing procedures that ensure the privacy
from data in order to better understand the situations of census information.
that these data portray. Statisticians improve data quality by developing
Scope of Statistics specific experiment designs and survey samples.
Some consider statistics a mathematical body of Statistics itself also provides tools for prediction
science that pertains to the collection, analysis, and forecasting the use of data and statistical

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 455
ISSN 2229-5518

models. Statistics is applicable to a wide variety of of computational statistics and the design of

academic disciplines, including natural and social experiments.

sciences, government, and business. Statistical Statistics is closely related to probability theory,

consultants can help organizations and companies with which it is often grouped. The difference is,
that don't have in-house expertise relevant to their roughly, that probability theory starts from the

particular questions. given parameters of a total population to deduce

Statistical methods can summarize or describe a probabilities that pertain to samples. Statistical

collection of data. This is called descriptive inference, however, moves in the opposite

statistics. This is particularly useful in direction—inductively inferring from samples to

communicating the results of experiments and the parameters of a larger or total population.

research. In addition, data patterns may be History of Statistics

modeled in a way that accounts for randomness Statistical methods date back at least to the 5th

and uncertainty in the observations. century BC. The earliest known writing on
These models can be used to draw inferences about statistics appears in a 9th-century book entitled

IJSER
the process or population under study—a practice Manuscript on Deciphering Cryptographic

called inferential statistics. Inference is a vital Messages, written by Al-Kindi. In this book, Al-
element of scientific advance, since it provides a Kindi provides a detailed description of how to use
way to draw conclusions from data that are subject statistics and frequency analysis to decipher

to random variation. To prove the propositions encrypted messages. This was the birth of both
being investigated further, the conclusions are statistics and cryptanalysis, according to the Saudi
tested as well, as part of the scientific method. engineer Ibrahim Al-Kadi.
Descriptive statistics and analysis of the new data The Nuova Cronica, a 14th-century history of
tend to provide more information as to the truth of Florence by the Florentine banker and official

the proposition. Giovanni Villani, includes much statistical

"Applied statistics" comprises descriptive statistics information on population, ordinances, commerce,

and the application of inferential statistics (Al- education, and religious facilities, and has been

Kadi, 1992). Theoretical statistics concerns both the described as the first introduction of statistics as a
logical arguments underlying justification of positive element in history (Singh, 2000).

approaches to statistical inference, as well Some scholars pinpoint the origin of statistics to

encompassing mathematical statistics. 1663, with the publication of Natural and Political
Mathematical statistics includes not only the Observations upon the Bills of Mortality by John

manipulation of probability distributions necessary Graunt. Early applications of statistical thinking

for deriving results related to methods of revolved around the needs of states to base policy
estimation and inference, but also various aspects on demographic and economic data, hence its stat-

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 456
ISSN 2229-5518

etymology. The scope of the discipline of statistics can then be subjected to statistical analysis, serving

broadened in the early 19th century to include the two related purposes: description and inference.
collection and analysis of data in general. Today, • Descriptive statistics summarize the

statistics is widely employed in government, population data by describing what was

business, and natural and social sciences. observed in the sample numerically or

Its mathematical foundations were laid in the 17th graphically. Numerical descriptors
century with the development of the probability include mean and standard deviation for

theory by Blaise Pascal and Pierre de Fermat. continuous data types (like heights or

Probability theory arose from the study of games weights), while frequency and percentage
of chance. The method of least squares was first are more useful in terms of describing

described by Adrien-Marie Legendre in 1805. The categorical data (like race).

use of modern computers has expedited large-scale • Inferential statistics uses patterns in the
statistical computation, and has also made possible sample data to draw inferences about the
new methods that are impractical to perform population represented, accounting for

IJSER
manually. randomness. These inferences may take
Statistical Overview the form of: answering yes/no questions
In applying statistics to a scientific, industrial, or about the data (hypothesis testing),
societal problem, it is necessary to begin with a estimating numerical characteristics of the
population or process to be studied. Populations data (estimation), describing associations
can be diverse topics such as "all persons living in within the data (correlation) and modeling
a country" or "every atom composing a crystal". A relationships within the data (for example,
population can also be composed of observations using regression analysis). Inference can
of a process at various times, with the data from extend to forecasting, prediction and
each observation serving as a different member of estimation of unobserved values either in
the overall group. Data collected about this kind of or associated with the population being
"population" constitutes what is called a time studied; it can include extrapolation and

series. interpolation of time series or spatial data,

For practical reasons, a chosen subset of the and can also include data mining (Willcox,
population called a sample is studied—as opposed 1938).
to compiling data about the entire group (an "... it is only the manipulation of uncertainty that
operation called census). Once a sample that is interests us. We are not concerned with the matter
representative of the population is determined, that is uncertain. Thus we do not study the
data is collected for the sample members in an mechanism of rain; only whether it will rain."
observational or experimental setting. This data

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 457
ISSN 2229-5518

According to Dennis Lindley, 2000, the concept of "statistical theory") to study the sampling
correlation is particularly noteworthy for the distributions of sample statistics and, more
potential confusion it can cause. Statistical analysis generally, the properties of statistical procedures.
of a data set often reveals that two variables The use of any statistical method is valid when the

(properties) of the population under consideration system or population under consideration satisfies

tend to vary together, as if they were connected. the assumptions of the method.
For example, a study of annual income that also Misuse of statistics can produce subtle, but serious
looks at age of death might find that poor people errors in description and interpretation—subtle in
tend to have shorter lives than affluent people. The the sense that even experienced professionals make
two variables are said to be correlated; however, such errors, and serious in the sense that they can
they may or may not be the cause of one another. lead to devastating decision errors. For instance,
The correlation phenomena could be caused by a social policy, medical practice, and the reliability of
third, previously unconsidered phenomenon, structures like bridges all rely on the proper use of
called a lurking variable or confounding variable. statistics.

IJSER
For this reason, there is no way to immediately Even when statistical techniques are correctly

infer the existence of a causal relationship between applied, the results can be difficult to interpret for
the two variables. those lacking expertise. The statistical significance
To use a sample as a guide to an entire population, of a trend in the data—which measures the extent
it is important that it truly represent the overall to which a trend could be caused by random
population. Representative sampling assures that variation in the sample—may or may not agree
inferences and conclusions can safely extend from with an intuitive sense of its significance. The set of
the sample to the population as a whole. A major basic statistical skills (and skepticism) that people
problem lies in determining the extent that the need to deal with information in their everyday
sample chosen is actually representative. Statistics lives properly is referred to as statistical literacy.

offers methods to estimate and correct for any Key terms used in statistics

random trending within the sample and data Null hypothesis: Interpretation of statistical

collection procedures. There are also methods of information can often involve the development of
experimental design for experiments that can a null hypothesis in that the assumption is that
lessen these issues at the outset of a study, whatever is proposed as a cause has no effect on
strengthening its capability to discern truths about the variable being measured.
the population. The best illustration for a novice is the predicament
Randomness is studied using the mathematical encountered by a jury trial. The null hypothesis,

discipline of probability theory. Probability is used H 0 , asserts that the defendant is innocent, whereas

in "mathematical statistics" (alternatively, the alternative hypothesis, H 1 , asserts that the

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 458
ISSN 2229-5518

defendant is guilty. The indictment comes because whole population. Any estimates obtained from
of suspicion of the guilt. The H 0 (status quo) stands the sample only approximate the population value.

in opposition to H 1 and is maintained unless H 1 is Confidence intervals allow statisticians to express

supported by evidence"beyond a reasonable how closely the sample estimate matches the true
doubt". However,"failure to reject H 0 " in this case value in the whole population. Often they are

does not imply innocence, but merely that the expressed as 95% confidence intervals. Formally, a

evidence was insufficient to convict. So the jury 95% confidence interval for a value is a range

does not necessarily accept H 0 but fails to reject H0 . where, if the sampling and analysis were repeated

While one cannot "prove" a null hypothesis one under the same conditions (yielding a different
can test how close it is to being true with a power dataset), the interval would include the true
test, which tests for type II errors (Thompson, (population) value 95% of the time. This does not
2006). imply that the probability that the true value is in
Error: Working from a null hypothesis two basic the confidence interval is 95%. From the frequentist
forms of error are recognized: perspective, such a claim does not even make

IJSER
• Type I errors where the null hypothesis is sense, as the true value is not a random variable.

falsely rejected giving a "false positive". Either the true value is or is not within the given
• Type II errors where the null hypothesis interval. However, it is true that, before any data
fails to be rejected and an actual difference are sampled and given a plan for how to construct

between populations is missed giving a the confidence interval, the probability is 95% that

false negative. the yet-to-be-calculated interval will cover the true

Error also refers to the extent to which individual value: at this point, the limits of the interval are

observations in a sample differ from a central yet-to-be-observed random variables. One

value, such as the sample or population mean. approach that does yield an interval that can be

Many statistical methods seek to minimize the interpreted as having a given probability of

mean-squared error, and these are called "methods containing the true value is to use a credible

of least squares." interval from Bayesian statistics: this approach

Measurement processes that generate statistical depends on a different way of interpreting what is

data are also subject to error. Many of these errors meant by "probability", that is as a Bayesian

are classified as random (noise) or systematic probability.

(bias), but other important types of errors (e.g., Significance: Statistics rarely give a simple Yes/No

blunder, such as when an analyst reports incorrect type answer to the question asked of them.

units) can also be important. Interpretation often comes down to the level of

Interval estimation: Most studies only sample part statistical significance applied to the numbers and

of a population, so results don't fully represent the often refers to the probability of a value accurately

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 459
ISSN 2229-5518

rejecting the null hypothesis (sometimes referred values, they describe both the size of the effect and

to as the p-value). the uncertainty surrounding it.

Referring to statistical significance does not Research Method used is the use of statistical tools

necessarily mean that the overall result is to analyze the production mix data, in order to

significant in real world terms. For example, in a understand what the data portrays.

large study of a drug it may be shown that the Table 1: Quantity of the raw
drug has a statistically significant but very small material for Soap Production Mix
Seria (X11) (X12 (X13 (X14) (X15 (Y)
beneficial effect, such that the drug is unlikely to
l No ) ) (liters ) (kg
help the patient noticeably. (liters (kg) (kg) ) (kg) )
)
Criticisms arise because the hypothesis testing
1 1500 26 50 150 69 360
approach forces one hypothesis (the null 2 1500 25 63 150 67 364
3 1500 25 64 175 69 365
hypothesis) to be "favored," and can also seem to
4 1500 25 62 150 66 363
exaggerate the importance of minor differences in 5 1500 28 68 175 68 366
large studies. A difference that is highly 6 1500 26 70 175 70 367

IJSER
7 1500 25 63 150 65 364
statistically significant can still be of no practical 8 1500 27 65 175 66 365
significance, but it is possible to properly formulate 9 1500 26 70 175 70 367
10 1500 28 68 175 70 366
tests in account for this. (See also criticism of 11 1500 25 62 150 67 363
hypothesis testing.) 12 1500 25 50 150 67 362
13 1500 28 70 175 68 367
One response involves going beyond reporting
14 1500 26 50 150 68 361
only the significance level to include the p-value 15 1500 28 68 175 67 366
16 1500 26 63 150 69 364
when reporting whether a hypothesis is rejected or
accepted. The p-value, however, does not indicate Where X11=qty of oil used (in liters), X12 = qyt of
the size of the effect. A better and increasingly salt used (in kg), X13 = Quantity of starch used (in
common approach is to report confidence kg), X14 = quantity of silicate used (in liters), X15 =
intervals. Although these are produced from the Total fatty matter in oil (in %), Y = Quantity of soap
same calculations as those of hypothesis tests or p- produced (in units of 11kg)

Table 2: Bivariate Correlations of Soap Production Mix Variables

X11 X12 X13 X14 X15 Y
a a a a a a
Pearson Correlation . . . . . .
Sig. (2-tailed) . . . . .
X11
Sum of Squares and Cross- .000 .000 .000 .000 .000 .000
products

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 460
ISSN 2229-5518

Covariance .000 .000 .000 .000 .000 .000

N 16 16 16 16 16 16
a ** *
Pearson Correlation . 1 .470 .686 .260 .528
Sig. (2-tailed) . .066 .003 .330 .036
Sum of Squares and Cross- .000 22.438 60.375 162.500 7.375 20.875
X12
products
Covariance .000 1.496 4.025 10.833 .492 1.392
N 16 16 16 16 16 16
a ** **
Pearson Correlation . .470 1 .737 .202 .952
Sig. (2-tailed) . .066 .001 .453 .000
Sum of Squares and Cross- .000 60.375 735.750 1000.000 32.750 215.750
X13
products
Covariance .000 4.025 49.050 66.667 2.183 14.383
N 16 16 16 16 16 16
a ** **
Pearson Correlation . .686 .737 1 .418 .838**
Sig. (2-tailed) . .003 .001 .107 .000

IJSER
Sum of Squares and Cross- .000 162.500 1000.000 2500.000 125.000 350.000
X14
products
Covariance .000 10.833 66.667 166.667 8.333 23.333
N 16 16 16 16 16 16
a
Pearson Correlation . .260 .202 .418 1 .295
Sig. (2-tailed) . .330 .453 .107 .267
Sum of Squares and Cross- .000 7.375 32.750 125.000 35.750 14.750
X15
products
Covariance .000 .492 2.183 8.333 2.383 .983
N 16 16 16 16 16 16
a * ** **
Pearson Correlation . .528 .952 .838 .295 1
Sig. (2-tailed) . .036 .000 .000 .267
Sum of Squares and Cross- .000 20.875 215.750 350.000 14.750 69.750
Y
products
Covariance .000 1.392 14.383 23.333 .983 4.650
N 16 16 16 16 16 16
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).

Table 3: Nonparametric Correlations of Soap Production Mix Variables

X11 X12 X13 X14 X15 Y
Kendall's tau_b X11 Correlation Coefficient . . . . . .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 461
ISSN 2229-5518

Sig. (2-tailed) . . . . . .
N 16 16 16 16 16 16
* **
Correlation Coefficient . 1.000 .461 .623 .281 .426*
X12 Sig. (2-tailed) . . .030 .009 .189 .043
N 16 16 16 16 16 16
. .461* 1.00 .773** .303 .982**
Correlation Coefficient
0
X13
Sig. (2-tailed) . .030 . .001 .135 .000
N 16 16 16 16 16 16
** ** **
Correlation Coefficient . .623 .773 1.000 .368 .766
X14 Sig. (2-tailed) . .009 .001 . .109 .001
N 16 16 16 16 16 16
Correlation Coefficient . .281 .303 .368 1.000 .282
X15 Sig. (2-tailed) . .189 .135 .109 . .162

IJSER
N 16 16 16 16 16 16
* ** **
Correlation Coefficient . .426 .982 .766 .282 1.000
Y Sig. (2-tailed) . .043 .000 .001 .162 .
N 16 16 16 16 16 16
Correlation Coefficient . . . . . .
X11 Sig. (2-tailed) . . . . . .
N 16 16 16 16 16 16
Correlation Coefficient . 1.000 .600* .670** .350 .568*
X12 Sig. (2-tailed) . . .014 .005 .184 .022
N 16 16 16 16 16 16
* **
. .600 1.00 .879 .408 .996**
Correlation Coefficient
0
X13
Sig. (2-tailed) . .014 . .000 .117 .000
Spearman's rho
N 16 16 16 16 16 16
** ** **
Correlation Coefficient . .670 .879 1.000 .414 .877
X14 Sig. (2-tailed) . .005 .000 . .111 .000
N 16 16 16 16 16 16
Correlation Coefficient . .350 .408 .414 1.000 .401
X15 Sig. (2-tailed) . .184 .117 .111 . .123
N 16 16 16 16 16 16
* ** **
Correlation Coefficient . .568 .996 .877 .401 1.000
Y
Sig. (2-tailed) . .022 .000 .000 .123 .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 462
ISSN 2229-5518

N 16 16 16 16 16 16
*. Correlation is significant at the 0.05 level (2-tailed).
**. Correlation is significant at the 0.01 level (2-tailed).

Table 4: T-Test of Soap Production Mix Variables

One-Sample Test
Test Value = 0
t df Sig. (2-tailed) Mean Difference 95% Confidence Interval of the
Difference
Lower Upper
X12 85.647 15 .000 26.18750 25.5358 26.8392
X13 35.910 15 .000 62.87500 59.1431 66.6069
X14 50.349 15 .000 162.50000 155.6208 169.3792

IJSER
X15 175.864 15 .000 67.87500 67.0524 68.6976
Y 675.899 15 .000 364.37500 363.2259 365.5241

Table 5: Partial Correlation of Soap Production Mix Variables

Correlations
Control Variables X11 X12 X13 X14 X15 Y
Correlation 1.000 . . . . .
X11 Significance (2-tailed) . . . . . .
df 0 14 14 14 14 14
Correlation . 1.000 .470 .686 .260 .528
X12 Significance (2-tailed) . . .066 .003 .330 .036
df 14 0 14 14 14 14
-none-a Correlation . .470 1.000 .737 .202 .952
X13 Significance (2-tailed) . .066 . .001 .453 .000
df 14 14 0 14 14 14
Correlation . .686 .737 1.000 .418 .838
X14 Significance (2-tailed) . .003 .001 . .107 .000
df 14 14 14 0 14 14
X15 Correlation . .260 .202 .418 1.000 .295

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 463
ISSN 2229-5518

Significance (2-tailed) . .330 .453 .107 . .267

df 14 14 14 14 0 14
Correlation . .528 .952 .838 .295 1.000
Y Significance (2-tailed) . .036 .000 .000 .267 .
df 14 14 14 14 14 0
Correlation 1.000 . . . .
X11 Significance (2-tailed) . . . . .
df 0 13 13 13 13
Correlation . 1.000 -.126 .526 .129
X12 Significance (2-tailed) . . .654 .044 .647
df 13 0 13 13 13
Correlation . -.126 1.000 -.366 -.273
Y X13 Significance (2-tailed) . .654 . .179 .326
df 13 13 0 13 13

IJSER
Correlation . .526 -.366 1.000 .327
X14 Significance (2-tailed) . .044 .179 . .234
df 13 13 13 0 13
Correlation . .129 -.273 .327 1.000
X15 Significance (2-tailed) . .647 .326 .234 .
df 13 13 13 13 0
a. Cells contain zero-order (Pearson) correlations.

Table 6: General Linear Model of Soap Production Mix Variables

Multivariate Testsa
Effect Value F Hypothesis df Error df Sig.
b
Pillai's Trace .996 243.971 1.000 1.000 .041
Wilks' Lambda .004 243.971b 1.000 1.000 .041
Intercept b
Hotelling's Trace 243.971 243.971 1.000 1.000 .041
Roy's Largest Root 243.971 243.971b 1.000 1.000 .041
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
X12 b
Wilks' Lambda 1.000 . .000 1.000 .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 464
ISSN 2229-5518

b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .916 3.623 3.000 1.000 .364
b
Wilks' Lambda .084 3.623 3.000 1.000 .364
X13 b
Hotelling's Trace 10.870 3.623 3.000 1.000 .364
b
Roy's Largest Root 10.870 3.623 3.000 1.000 .364
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X14 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .974 7.622 5.000 1.000 .268
b
Wilks' Lambda .026 7.622 5.000 1.000 .268
X15 b
Hotelling's Trace 38.111 7.622 5.000 1.000 .268
b
Roy's Largest Root 38.111 7.622 5.000 1.000 .268
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .

IJSER
X11 * X12 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X13 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X14 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X12 * X13 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X12 * X14 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
X12 * X15 b
Wilks' Lambda 1.000 . .000 1.000 .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 465
ISSN 2229-5518

b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X13 * X14 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X13 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X14 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .

IJSER
X11 * X12 * X13 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X12 * X14 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X12 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X13 * X14 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X13 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X11 * X14 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
X12 * X13 * X14 b
Wilks' Lambda 1.000 . .000 1.000 .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 466
ISSN 2229-5518

b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X12 * X13 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X12 * X14 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
Wilks' Lambda 1.000 . .000 1.000 .
X13 * X14 * X15 b
Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
X11 * X12 * X13 * Wilks' Lambda 1.000 . .000 1.000 .

IJSER
b
X14 Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
X11 * X12 * X13 * Wilks' Lambda 1.000 . .000 1.000 .
b
X15 Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
X11 * X12 * X14 * Wilks' Lambda 1.000 . .000 1.000 .
b
X15 Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
X11 * X13 * X14 * Wilks' Lambda 1.000 . .000 1.000 .
b
X15 Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
X12 * X13 * X14 * Wilks' Lambda 1.000 . .000 1.000 .
b
X15 Hotelling's Trace .000 . .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .
b
Pillai's Trace .000 . .000 .000 .
b
X11 * X12 * X13 * Wilks' Lambda 1.000 . .000 1.000 .
X14 * X15 Hotelling's Trace .000 .b .000 2.000 .
b
Roy's Largest Root .000 .000 1.000 .000 .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 467
ISSN 2229-5518

a. Design: Intercept + X11 + X12 + X13 + X14 + X15 + X11 * X12 + X11 * X13 + X11 * X14 + X11 * X15 +
X12 * X13 + X12 * X14 + X12 * X15 + X13 * X14 + X13 * X15 + X14 * X15 + X11 * X12 * X13 + X11 * X12 *
X14 + X11 * X12 * X15 + X11 * X13 * X14 + X11 * X13 * X15 + X11 * X14 * X15 + X12 * X13 * X14 + X12 *
X13 * X15 + X12 * X14 * X15 + X13 * X14 * X15 + X11 * X12 * X13 * X14 + X11 * X12 * X13 * X15 + X11 *
X12 * X14 * X15 + X11 * X13 * X14 * X15 + X12 * X13 * X14 * X15 + X11 * X12 * X13 * X14 * X15
b. Exact statistic

Table 7: Statistical Analysis for the Modeling of the Soap Production Mix Variables
Tests of Between-Subjects Effects
Source Dependent Variable Type III Sum of df Mean Square F Sig.
Squares
a
Y 69.750 14 4.982 . .
Corrected Model b
VAR00012 335.500 14 23.964 5.325 .329
Y 1859691.713 1 1859691.713 . .
Intercept
VAR00012 1097.869 1 1097.869 243.971 .041

IJSER
Y .000 0 . . .
X11
VAR00012 .000 0 . . .
Y .000 0 . . .
X12
VAR00012 .000 0 . . .
Y 9.500 3 3.167 . .
X13
VAR00012 48.917 3 16.306 3.623 .364
Y .000 0 . . .
X14
VAR00012 .000 0 . . .
Y .500 5 .100 . .
X15
VAR00012 171.500 5 34.300 7.622 .268
Y .000 0 . . .
X11 * X12
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X13
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X13
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X15
VAR00012 .000 0 . . .

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 468
ISSN 2229-5518

Y .000 0 . . .
X13 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X13 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X13
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X13 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X13 * X15
VAR00012 .000 0 . . .

IJSER
Y .000 0 . . .
X11 * X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X13 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X13 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X13 * X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X13 * X14
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X13 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X13 * X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X12 * X13 * X14 * X15
VAR00012 .000 0 . . .
Y .000 0 . . .
X11 * X12 * X13 * X14 * X15
VAR00012 .000 0 . . .
Y .000 1 .000
Error
VAR00012 4.500 1 4.500

IJSER © 2014
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 469
ISSN 2229-5518

Y 2124376.000 16
Total
VAR00012 1496.000 16
Y 69.750 15
Corrected Total
VAR00012 340.000 15
a. R Squared = 1.000 (Adjusted R Squared = 1.000)
b. R Squared = .987 (Adjusted R Squared = .801)

Discussion and Conclusion 3. Hays, William Lee, (1973) Statistics for the

From the analysis above, the results show the statistical Social Sciences, Holt, Rinehart and
analysis of the production mix data. This T test showed Winston, p.xii, ISBN 978-0-03-077945-9
that the data is significance to model and to analyze. 4. Moore, David (1992). "Teaching Statistics
The correlation analysis showed the impact of the
as a Respectable Subject". In F. Gordon
variables relationships in each other. Both partial
and S. Gordon. Statistics for the Twenty-
correlation and bivariate correlation show that X11(i.e.
First Century. Washington, DC: The
qty of oil used), is constant and it cannot correlate with

IJSER
other variables. The results also showed that only Mathematical Association of America.
X15(i.e. Total fatty matter in oil in %) is not significant pp. 14–25. ISBN 978-0-88385-078-7.
to the correlation of the other variables. However, the 5. Chance, Beth L.; Rossman, Allan J. (2005).
design model developed can be used in redesigning the
"Preface". Investigating Statistical
production mix of the soap production in the case study
Concepts, Applications, and Methods.
industry. The coefficient of determination (R Squared =
.987) produced by the model shows that the model is fit Duxbury Press. ISBN 978-0-495-05064-3.

for design of the production mix in the case study 6. Anderson, D.R.; Sweeney, D.J.; Williams,
company. T.A.. (1994) Introduction to Statistics:
In conclusion, it is observed that there is a need to test Concepts and Applications, pp. 5–9. West
and to analyze the data statistically in order to
Group. ISBN 978-0-314-03309-3
understand what the data portrays. The researcher has
7. Al-Kadi, Ibrahim A. (1992) "The origins of
recommended the design model to the case study
industry for wider use and applicability in there cryptology: The Arab contributions”,

company. Cryptologia, 16(2) 97–126.

References doi:10.1080/0161-119291866801

1. Dodge, Y. (2006) The Oxford Dictionary of 8. Singh, Simon (2000). The code book : the

Statistical Terms, OUP. ISBN 0-19-920613- science of secrecy from ancient Egypt to
9 quantum cryptography (1st Anchor Books

2. Moses, Lincoln E. (1986) Think and ed.). New York: Anchor Books. ISBN 0-
Explain with Statistics, Addison-Wesley, 385-49532-3.[page needed]
ISBN 978-0-201-15619-5 . pp. 1–3

IJSER © 2014
http://www.ijser.org
 International Journal of Scientific & Engineering Research, Volume 5, Issue 12, December-2014 470
ISSN 2229-5518

9. ^ Villani, Giovanni. Encyclopædia

Britannica. Encyclopædia Britannica 2006

Ultimate Reference Suite DVD. Retrieved

on 2008-03-04.
10. Willcox, Walter (1938) "The Founder of

Statistics". Review of the International

Statistical Institute 5(4):321–328.

JSTOR 1400906

11. Breiman, Leo (2001). "Statistical

Modelling: the two cultures". Statistical
Science 16 (3): 199–231.

doi:10.1214/ss/1009213726. MR 1874152.
CiteSeerX: 10.1.1.156.4933.
12. Lindley, D. (2000). "The Philosophy of

IJSER
Statistics". Journal of the Royal Statistical

Society, Series D 49 (3): 293–337.

doi:10.1111/1467-9884.00238.
JSTOR 2681060.
13. Thompson, B. (2006). Foundations of

behavioral statistics. New York, NY:

Guilford Press.

IJSER © 2014
http://www.ijser.org

View publication stats