HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY

OFFICE FOR INTERNATIONAL STUDY PROGRAMS

REPORT ASSIGNMENT

PROBABILITY AND STATISTICS

Lecturer: Mr. NGUYEN TIEN DUNG

ClassCC01-Group 6
No. Name Student ID
1 Nguyễn Anh Hào 1852019
School
2 Trịnh Minh Tiến 1752544 year:
3 Nguyễn Đô Trưởng 1752586
STATISTICAL METHOD 2020
DESCRIPTIVE STATISTICS (2 TESTS)
FREQUENTIST STATISTIC (4 TESTS)

CONTENTS

1|P a g e
I. Introduction: .........................................................................................................3

II. Descriptive statistics............................................................................................7

1. Q-Q plots for checking normality.................................................................7

2. Histogram with Normal Curve Overlay........................................................9

III. Frequentist statistics:..........................................................................................11

1. Introduction to the hypothesis testing...........................................................11

2. Z-Test method...............................................................................................14

3. Combined “Lilliefors corrected” K-s test method........................................19

4. Shapiro-Wilk original test method................................................................22

5. Anderson-Darling test method......................................................................26

IV. Summary & references:....................................................................................30

2|P a g e
I. Introduction

Analyze the data:

For the project of probability and statistics, we want to evaluate our data, through statistic
process to determine whether it is normally distributed.
So what is “normally distributed”? – In statistics, normally distributed, also known as the
Gaussian distribution, is a probability distribution that is symmetric about the mean,
showing that data near the more frequent in occurrence than data far from the mean. In
graph form, normal distribution will appear as a bell curve.
To testing the data set is well-modeled by a normal distribution and to compute how
likely it is for a random variable underlying the data set to be normally distributed.

+ In descriptive statistics terms, one measures a goodness of fit of a normal model to the
data – if the fit is poor then the data are not well modeled in that respect by a normal
distribution, without making a judgment on any underlying variable.

+ In frequentist statistics statistical hypothesis testing, data are tested against the null
hypothesis that it is normally distributed.

+ In Bayesian statistics, one does not "test normality" per se, but rather computes the
likelihood that the data come from a normal distribution with given parameters μ,σ (for
all μ,σ), and compares that with the likelihood that the data come from other distributions
under consideration, most simply using a Bayes factor (giving the relative likelihood of
seeing the data given different models), or more finely taking a prior distribution on
possible models and parameters and computing a posterior distribution given the
computed likelihoods.
For our specific project, we decided to use descriptive statistics and Frequentist statistics
approach to our data set.

3|P a g e
DATA SET: CORE FREQUENCY RECORDED VALUE:

The sample size may vary depends on the pc components (of which are replaceable and
convertible). Therefore, the possible population could be massive (might altering to
countless). In this case, we only pick the phenomenal individuals to put in the sample (as
they have the complete data set of all components and features).

ID VALUE ID VALUE ID VALUE

1 5.4 18 5.1 35 5
2 5.3 19 5.1 36 5
3 5.2 20 5.1 37 5
4 5.2 21 5.1 38 5
5 5.2 22 5.005 39 4.954
6 5.2 23 5 40 4.9
7 5.2 24 5 41 4.9
8 5.2 25 5 42 4.9
9 5.2 26 5 43 4.9
10 5.2 27 5 44 4.9
11 5.197 28 5 45 4.9
12 5.125 29 5 46 4.8
13 5.1 30 5 47 4.8
14 5.1 31 5 48 4.8
15 5.1 32 5 49 4.72
16 5.1 33 5 50 4.7
17 5.1 34 5

4|P a g e
REPAIRING BEFORE CODING:
 Packages and libraries need to be installed before computing with R code via
Rstudio.
CODE EXPLAINATION
>[Link](“dplyr” “dplyr” is the next iteration of “plyr”, focussed on
) tools for working with data frames (hence the d in
the name). This action will install the latest released
> library(dplyr) version from CRAN

It has three main goals:

• Identify the most important data manipulation tools needed for data analysis and
make them easy to use from R.
• Provide blazing fast performance for in-memory data by writing key pieces in C+
+.
• Use the same interface to work with data no matter where it's stored, whether in a
data frame, a data table or database.
Attaching package: ‘dplyr’

The following objects are masked from ‘package:stats’:

filter, lag

The following objects are masked from ‘package:base’:

intersect, setdiff, setequal, union

CODE EXPLAINATION
>[Link]("ggpl “ggplot2” is a system for declaratively creating
ot2") graphics, based on The Grammar of Graphics. You
provide the data, tell “ggplot2” how to map variables
>library(ggplot2) to aesthetics, what graphical primitives to use, and it
takes care of the details.

>library(nortest) Five omnibus tests for testing the composite

hypothesis of normality.

5|P a g e
 CODE EXPLAINATION
“EnvStats” is Package for
Environmental Statistics, Including
> library(EnvStats) US EPA Guidance.

Graphical and statistical analyses of

environmental data, with focus on
analyzing chemical concentrations
and physical parameters, usually in
the context of mandated
environmental monitoring.

Major environmental statistical

methods found in the literature and
regulatory guidance documents,
with extensive help that explains
what these methods do, how to use
them, and where to find them in the
literature.

Numerous built-in data sets from

regulatory guidance documents and
environmental statistics literature.
> O.C_RECORD=c( 5.4,5.3,5.2,5.2,5.2,5.2,5.2,5.2,
5.2,5.2,5.197,5.125,5.1,5.1,5.1,5.1,5.1,5.1,5.1,5.1,
5.1,5.005,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,4.954,4.9, To create a dataset, we use the
4.9,4.9,4.9,4.9,4.9,4.8,4.8,4.8,4.72,4.7) following code line.

To create QQplot, use the following

> qqnorm(O.C_RECORD) code line. Note: the “qqnorm” line
> qqline(O.C_RECORD) must be written before the “qqline”
as if the R required the y axel
before the x axel.
> RECORD_2017=[Link](O.C_RECORD) To create the table, use the
> View(RECORD_2017) following code line.

6|P a g e
II. Descriptive statistics

1. Q-Q Plots for checking normality

A. Theory

The Q-Q plot, or quantile-quantile plot, is a graphical tool to help us assess if a set of
data plausibly came from some theoretical distribution such as a Normal or exponential.

A Q-Q plot is a scatterplot created by plotting two sets of quantiles against one
another. If both sets of quantiles came from the same distribution, we should see the
points forming a line that’s roughly straight. Here’s an example of a Normal Q-Q plot
when both sets of quantiles truly come from Normal distributions.

In R, there are two functions to create Q-Q plots: “qqnorm” and “qqpline”.
 The “qqnorm” creates a Normal Q-Q plot. You give it a vector of data and R plots
the data in sorted order versus quantiles from a standard Normal distribution.
 The “qqpline” function allows you to create a Q-Q plot for any distribution. Unlike
the qqnorm function, you have to provide two arguments: the first set of data and the
second set of data.

Notice the points fall along a line in the middle of the graph, but curve off in the
extremities. Normal Q-Q plots that exhibit this behavior usually mean your data have
more extreme values than would be expected if they truly came from a Normal
distribution.

7|P a g e
B. Code

>
O.C_RECORD=c(5.4,5.3,5.2,5.2,5.2,5.2,5.2,5.2,5.2,5.2,5.
197,5.125,5.1,5.1,5.1,5.1,5.1,5.1,5.1,5.1,5.1,5.005,5,5
,5,5,5,5,5,5,5,5,5,5,5,5,5,5,4.954,4.9,4.9,4.9,4.9,4.9,
4.9,4.8,4.8,4.8,4.72,4.7)
> RECORD_2017=[Link](O.C_RECORD)
> View(RECORD_2017)
> View(RECORD_2017)
> qqnorm(O.C_RECORD)
> qqline(O.C_RECORD)

2. Histogram with Normal Curve Overlay

A. Theory

 A histogram is an approximate representation of the distribution of numerical or

categorical data. It was first introduced by Karl Pearson.

 To construct a histogram, the first step is to "bin" (or "bucket") the range of values,
divide the entire range of values into a series of intervals and then count how many
values fall into each interval. The bins are usually specified as consecutive, non-
overlapping intervals of a variable. The bins (intervals) must be adjacent, and are often
(but not required to be) of equal size.

8|P a g e
 If the bins are of equal size, a rectangle is erected over the bin with height
proportional to the frequency, the number of cases in each bin.

 A histogram may also be normalized to display "relative" frequencies. It then

shows the proportion of cases that fall into each of several categories, with the sum of the
heights equaling.

 Histograms give a rough sense of the density of the underlying distribution of the
data, and often for density estimation: estimating the probability density function of the
underlying variable.

[Link]

9|P a g e
III. Frequentist statistics
1. Introduction to the hypothesis testing:
Definition:

The Hypothesis Testing is a statistical test used to determine whether the hypothesis
assumed for the sample of data stands true for the entire population or not. Simply, the
hypothesis is an assumption which is tested to determine the relationship between two
data sets.

In hypothesis testing, two opposing hypotheses about a population are formed. Null
Hypothesis (H0) and Alternative Hypothesis (H1). The Null hypothesis is the statement
which asserts that there is no difference between the sample statistic and population
parameter and is the one which is tested, while the alternative hypothesis is the statement
which stands true if the null hypothesis is rejected.

The following Hypothesis Testing Procedure is followed to test the assumption made:

1. Set up a Hypothesis: The first step is to establish the hypothesis to be tested. The
statistical hypothesis is an assumption about the value of some unknown parameter, and
the hypothesis provides some numerical value or range of values for the parameter. Here
two hypotheses about the population are constructed Null Hypothesis and Alternative
Hypothesis.

The Null Hypothesis denoted by H0 asserts that there is no true difference between the
sample of data and the population parameter and that the difference is accidental which is
caused due to the fluctuations in sampling. Thus, a null hypothesis states that there is no
difference between the assumed and actual value of the parameter.

The alternative hypothesis denoted by H1 is the other hypothesis about the population,
which stands true if the null hypothesis is rejected. Thus, if we reject H 0 then the
alternative hypothesis H1 gets accepted.

 For our frequentist statistics, our hypothesis:

Ho: The data follows the normal distribution.

10 | P a g e
H1: The data do not follow the normal distribution.

2. Set up a Suitable Significance Level: Once the hypothesis about the population is
constructed the researcher has to decide the level of significance, i.e. a confidence level
with which the null hypothesis is accepted or rejected. The significance level is denoted
by ‘α’ and is usually defined before the samples are drawn such that results obtained do
not influence the choice. In practice, we either take 5% or 1% level of significance.

If the 5% level of significance is taken, it means that there are five chances out of 100
that we will reject the null hypothesis when it should have been accepted, i.e. we are
about 95% confident that we have made the right decision. Similarly, if the 1% level of
significance is taken, it means that there is only one chance out of 100 that we reject the
hypothesis when it should have been accepted, and we are about 99% confident that the
decision made is correct.

 For our project, the confident level selected is 95%.

3. Determining a Suitable Test Statistic: After the hypothesis are constructed, and the
significance level is decided upon, the next step is to determine a suitable test statistic and
its distribution. Most of the statistic tests assume the following form:

Sample statistic−Hypothesized Parameter

Test statistic = Standard Error of the Statistic

4. Determining the Critical Region: Before the samples are drawn it must be decided
that which values to the test statistic will lead to the acceptance of H 0 and which will
lead to its rejection. The values that lead to rejection of H 0 is called the critical region.

5. Performing Computations: Once the critical region is identified, we compute several

values for the random sample of size ‘n.’ Then we will apply the formula of the test
statistic as shown in step (3) to check whether the sample results falls in the acceptance
region or the rejection region.

6. Decision-making: Once all the steps are performed, the statistical conclusions can be
drawn, and the management can take decisions. The decision involves either accepting
the null hypothesis or rejecting it. The decision that the null hypothesis is accepted or
rejected depends on whether the computed value falls in the acceptance region or the
rejection region.

11 | P a g e
Thus, to test the hypothesis, it is necessary to follow these steps systematically so that the
results obtained are accurate and do not suffer from either of the statistical error.

While testing the hypothesis, an individual may commit the

following types of error:

1. Type-I Error: True Null hypothesis is rejected, i.e. hypothesis is rejected when it
should be accepted. The probability of committing the type-I error is denoted by α and is
called as a level of significance.

If, α = Pr[type-I error] = Pr [reject H0/H0 is true]

Then, (1-α) = Pr[accept H0/H0 is true]
(1-α) = corresponds to the concept of Confidence Interval.

2. Type-II Error: A False Null hypothesis is accepted, i.e. hypothesis is accepted when
it should be rejected. The probability of committing the type-II error is denoted by β.

If, β = Pr[type-II error] = Pr[accept H0/H0 is false]

Then, (1-β) = Pr[reject Ho/H0 is false
(1-β) = power of a statistical test.

Thus, hypothesis testing is the important method in the statistical inference that measures
the deviations in the sample data from the population parameter. The hypothesis tests are
widely used in the business and industry for making the crucial business decisions.

12 | P a g e
2. Z-TEST method:
A. Theory & Calculation:

Theory:

The one-sample z-test is used to test whether the mean of a population is greater than,
less than, or not equal to a specific value. Because the standard normal distribution is
used to calculate critical values for the test, this test is often called the one-sample z-test.

The z-test assumes that the population standard deviation is known.

In addition, according to Khan Academy we can estimate our population standard

deviation using our sample standard deviation if our sample size is greater than 30 which
is in our case (n = 50).

We also use “BSDA”: Functions and data sets for the text Basic Statistics and Data
Analysis:

-->You can check everything is installed correctly with the has_devel() function from the
devtools package. Type the following at the R prompt:
[Link]("devtools", dependencies = TRUE)
devtools::has_devel()

-->If everything is installed correctly, the function will print some output and then return
TRUE.

-->To install the BSDA package, type the following at the R prompt:
devtools::install_github('alanarnholt/BSDA')

-->It is possible to install BSDA with GIT and the R CMD build assuming you have GIT
installed and the appropriate tools to build R from source.
git clone [Link]
R CMD build BSDA
R CMD INSTALL BSDA_*.[Link]

13 | P a g e
Purpose:

_Determine if there is any significant difference between the mean of population (core
frequency) and the hypothesized mean.
_Determine whether the population mean, the core frequency, is greater than 5.
_So, why is 5?

 Our data is recorded in 2017, in that year, the lowest standard to be considered a
successful overclock for a core frequency of a chip whilst overclocking is 5 GHz.
 We want to check if any setups in this list; which are recorded in different
motherboard, RAM, cooler, manufactured countries, …; reach that requirement (greater
than 5 GHz)
 Actually, there are not any significant different between 4.7, 4.8 and 5 GHz, this is
just something we really interested in gaming and basically, optimize and get the most
out of the chip with these setups.
 Moreover, for example, we want to build a PC for gaming. The different between
60 fps and 120 fps is actually easy to recognize, but there is not much noticeable point
between 90 fps and 120 fps. So that the reason why we need at least 5 GHz for “up to or
higher 90 fps” gaming experience with the purpose of economizing and performing
sufficiency.

Hypothesis:
The null hypothesis (H0) and alternative hypothesis (H1) of the one sample Z test can be
expressed as:
H0: x ≤ µ ("the sample mean is less or equal to the [proposed] population mean ")
H1: x > µ ("the sample mean is greater than the [proposed] population mean")
where µ is a constant proposed for the population mean and x is the sample mean.
Test Statistic:

_The test statistic for a One Sample Z-Test is denoted z, which is calculated using the
following formula:
_The sample mean:

14 | P a g e
 n

∑ X i = 5.03402
i=1
x=
n

_The test statistic

x−μ
z=
S /√n
=¿ 1.678

Where

-->We have the Z value is equal to 1.678 and the Z α =1.64 value based on Cumulative
Standard Normal Distribution table. With Z is greater than Z α so we reject the null
hypothesis that the sample mean is less or equal to 5.0.

-->In addition, we also look up to the Cumulative Standard Normal Distribution table and
use interpolating to get the result p = 0.00467.

15 | P a g e
We also use Excel for double-checking our result.

[Link]:

16 | P a g e
 CODE EXPLAINATION
To perform the Z test , use the following
code line.
> [Link](x=O.C_RECORD,
alternative="greater", mu=5.0, Note: The alternative could be changed
sigma.x = sd(O.C_RECORD), into “less” or “[Link]” depends on the
[Link] = 0.95) hypothesis theories being tested.
“Confident level” and “mu” may vary by
theories.

> [Link](x=O.C_RECORD, alternative="greater", mu=5.0, sigma.x =

sd(O.C_RECORD), [Link] = 0.95)
One-sample z-Test

data: O.C_RECORD
z = 1.678, p-value = 0.04668
alternative hypothesis: true mean is greater than 5
95 percent confidence interval:
5.000671 NA
sample estimates:
mean of x
5.034

17 | P a g e
 Conclusion:
In conclusion, we have the p-value equal to 0.0467 and the value is less than α (0.05)
so we reject the null hypothesis and concluding that the sample mean is greater than
5.0.
[Link] “LILLIEFORS CORRECTED” K-S TEST
method
A. Theory:
+ In statistics, the Lilliefors test is a normality test based on the Kolmogorov–Smirnov
test.
+ It is used to test the null hypothesis that data come from a normally
distributed population, when the null hypothesis does not specify which normal
distribution.
+ It does not specify the expected value and variance of the distribution and it is named
after Hubert Lilliefors, professor of statistics at George Washington University.
Process:
1. First estimate the population mean and population variance based on the data.
2. Then find the maximum discrepancy between the empirical distribution function and
the cumulative distribution function (CDF) of the normal distribution with the estimated
mean and estimated variance. Just as in the Kolmogorov–Smirnov test, this will be the
test statistic.
3. Finally, assess whether the maximum discrepancy is large enough to be statistically
significant, thus requiring rejection of the null hypothesis. This is where this test becomes
more complicated than the Kolmogorov–Smirnov test. Since the hypothesized CDF has
been moved closer to the data by estimation based on those data, the maximum
discrepancy has been made smaller than it would have been if the null hypothesis had
singled out just one normal distribution. Thus the "null distribution" of the test statistic,
i.e. its probability distribution assuming the null hypothesis is true, is stochastically
smaller than the Kolmogorov–Smirnov distribution.

Illustration of the Kolmogorov–

Smirnov statistic. Red line is CDF,
blue line is an ECDF, and the black

18 | P a g e
arrow is the K–S statistic.
Hypotheses:
The two hypotheses for the Lilliefors test for the normal distribution are given below:
 The null hypothesis (H0): The data follows the normal distribution.
 The alternative hypothesis (H1): The data do not follow the normal distribution.

 The general steps that the test follows are:

Calculate Zi using this formula:

Xi− X
Zi =
s

Where:
 Zi = the individual Z-Scores for every member in your sample.
 Xi = individual member / data point.
 X = sample mean.

Calculate the test statistic, which is the empirical distribution function (EDF) based
on the Zis. The formula is:

T 1= ¿|F∗( x ) −S (x)|
x

Where:
F*(x) = the standard normal distribution function.
S(x) = the empirical distribution function of the Zi values.

19 | P a g e
B. Code:
CODE EXPLAINATION
>[Link](O.C_RECORD To perform the Lilliefors test , use the following
) code line.

> [Link](O.C_RECORD)

Lilliefors (Kolmogorov-Smirnov) normality test

data: O.C_RECORD
D = 0.16621, p-value = 0.001418

Conclusion:
In conclusion, we have the p-value equal to 0.001418 and p-value < α (0.05) so we
reject the null hypothesis which show that the data do not follow the normal distribution.

[Link]-WILK ORIGINAL TEST method

A. Theory & Calculation:
Shapiro-Wilk Original Test
The Shapiro-Wilk test is a way to tell if a random sample comes from a normal
distribution. The test gives you a W value; small values of p indicate your sample is not
normally distributed (you can reject the null hypothesis that your population is normally
distributed if your values are under a certain threshold).
Hypotheses:
The two hypotheses for the Shapiro-Wilk test for the normal distribution are given below:
 The null hypothesis (H0): The data follows the normal distribution.
 The alternative hypothesis (H1): The data do not follow the normal distribution.

We also use Excel for double-checking our result.

The basic approach used in the Shapiro-Wilk (SW) test for normality is as follows:

20 | P a g e
 Rearrange the data in ascending order so that x1 ≤ … ≤ xn.
 Calculate SS as follows:
n
SS=∑ ( x i−x )2=1.00710698
i=1

 If n is even, let m = n/2, while if n is odd let m = (n–1)/2

 Calculate b as follows, taking the ai weights from the Table 1 (based on the value
of n) in the Shapiro-Wilk Tables. Note that if n is odd, the median data value is not used
in the calculation of b.

m
b=∑ a i (x n+1−i−x i)=0.9794221
i=1

After using the SW table, we have a full table of a i:

21 | P a g e
Calculate a i (x n+1−i−x i)

Calculate b:

2
b
Calculate the test statistic: W = =0.952498264
SS

Is made of the expected values of the order statistics of independent and identically
distributed random variables sampled from the standard normal distribution; Finally, V is
the covariance matrix of those normal order statistics. There is no name for the

22 | P a g e
distribution of W. The cutoff values for the statistics are calculated through Monte-Carlo
simulations.

Extra: [Link] (the link of S-W

table)

B. Code:
CODE EXPLAINATION
> [Link](O.C_RECORD) To perform the Shapiro-Wilk test , use the
following code line.

> [Link](O.C_RECORD)

Shapiro-Wilk normality test

data: O.C_RECORD
W = 0.95371, p-value = 0.04857
Conclusion:
In conclusion, we have the p-value equal to 0.04857 and p-value < α (0.05) so we
reject the null hypothesis and concluding that the data do not follow the normal
distribution.

[Link]-DARLING TEST method

A. Theory:

23 | P a g e
The Anderson–Darling test is a statistical test of whether a given sample of data is
drawn from a given probability distribution. In its basic form, the test assumes that there
are no parameters to be estimated in the distribution being tested, in which case the test
and its set of critical values is distribution-free. However, the test is most often used in
contexts where a family of distributions is being tested, in which case the parameters of
that family need to be estimated and account must be taken of this in adjusting either the
test-statistic or its critical values. When applied to testing whether a normal
distribution adequately describes a set of data, it is one of the most powerful statistical
tools for detecting most departures from normality.

The Anderson–Darling test assesses whether a sample comes from a specified

distribution. It makes use of the fact that, when given a hypothesized underlying
distribution and assuming the data does arise from this distribution, the cumulative
distribution function (CDF) of the data can be assumed to follow a uniform distribution.
The data can be then tested for uniformity with a distance test (Shapiro 1980). The
formula for the test statistic A to assess if data { Y 1 <…< Y n }

Hypotheses:
The two hypotheses for the Anderson-Darling test for the normal distribution are given
below:
• The null hypothesis (H0): The data follows the normal distribution.
• The alternative hypothesis (H1): The data do not follow the normal distribution.
(note that the data must be put in order) comes from a CDF is:
2
A =−n−S  A2=1.152581638

Where
n
2 i−1
S=∑
i=1 n [ ]
ln ( F ( Y i ) ) +ln ( 1−F ( Y n +1−i ) ) =−51.15258164

The test statistic can then be compared against the critical values of the theoretical
distribution. Note that in this case no parameters are estimated in relation to the
cumulative distribution function F.

24 | P a g e
We also use Excel for double-checking our result.
First, we calculate F(x) and F(x*) with the “mean” and “standard deviation”

then, calculate the “A” value,“S” value and “AD” value with the following formula:

(
AD= A 1+
0.75 2.25
n )
+ 2 =1.170907686
n

Then, we determine the p-value:

AD statistic P-Value Formula
2

A ≥ 0.6 p=e(1.2937 – 5.709 A +0.0186 A )

2

0.34 < AD < 0.6 p=e(0.9177 – 4.279 A +1.38 A )

2

0.2 < AD < 0.34 p=1−e(−8.318 – 42.796 A −59.938 A )

2

A≤ 0.2 < AD < 0.34 p=1−e(−13.436 – 101.14 A −223.73 A )

−→ P value=0.0046750710

25 | P a g e
The Anderson-Darling test can be used to answer the following questions:
 Are the data from a normal distribution or from a log-normal distribution?
 Are the data from a Weibull distribution?
 Are the data from an exponential distribution?
 Are the data from a logistic distribution?

Many statistical tests and procedures are based on specific distributional assumptions.
The assumption of normality is particularly common in classical statistical tests. Much
reliability modeling is based on the assumption that the data follow a Weibull
distribution.

There are many non-parametric and robust techniques that do not make strong
distributional assumptions. However, techniques based on specific distributional
assumptions are in general more powerful than non-parametric and robust techniques.
Therefore, if the distributional assumptions can be validated, they are generally preferred.

B. Code:
CODE EXPLAINATION

26 | P a g e
 >[Link](O.C_RECORD) To perform the Anderson-Darling test , use the
following code line.

> [Link](O.C_RECORD)

Anderson-Darling normality test

data: O.C_RECORD
A = 1.1526, p-value = 0.004675

Conclusion:
 In conclusion, we have the p-value equal to 0.004675 and p-value < α (0.05) so we
reject the null hypothesis and concluding that the data do not follow the normal
distribution.

IV. SUMMARY & REFERENCES:

Summary:
There are some slightly differences in the value because Excel has its limitations
regarding advanced statistics and calculations, quality and version control, user
experience and scalability. Switching to a more sophisticated data analysis tool or a
dashboard is often an answer. Transferring your Excel sheet to a R app can be the easiest
way to create an enterprise ready dashboard.

 [Link]
 [Link]

27 | P a g e
 tests-normality-symmetry/lilliefors-test-normality/
 [Link]

28 | P a g e