Page 1 of 117 FINAL REPORT OF BQT Authors Detail: This Report of BQT project is written by the Following persons at the end of the 5 th semester… Name: Raheel Tariq ID: 08108048 Student of BBA horns in Gift University Email ID: [email protected]Name: Aisha Riaz ID: 08108042 Student of BBA horns in Gift University Email ID: [email protected]Name: Komal Afzal ID: 08108034 Student of BBA horns in Gift University Email ID: [email protected]
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1 of 78 FINAL REPORT OF BQT
Authors Detail:
This Report of BQT project is written by the Following persons at the end of the 5th semester…
First of all we want to register our thanks to Prof. Sir. Abid Awan. Who opened our mind and shaped our future and have always been a friend, guide and philosopher to us. It was his inspiration and unbounded zeal that has created this project
Also thanks to Mr. Sarmad Hasan Territory sales Supervisor of Telenor, Mr. Tanweer Hussain Specialist customer care of Jazz and Mr. Usman Ahmed Regional Coordinator of Ufone who has given us enormous material support to our Project. It was quite pleasant for us that such nice people still exist in this world of sins. He gave us his precious time and briefed us personally during the visit.
We find an opportunity while writing this page to say thanks to our parents whom sleepless nights enabled us to accomplish this project. We always pray Allah for their good health. Their Shadow always remains on our head (Amen).
Last but not least, we want to register our thanks to our class fellows for being so cooperative, uncomplaining and no interrupting. It’s their matchless cooperation which enables us to study in the class peacefully.
Page 3 of 78 FINAL REPORT OF BQT
Objective:
The purpose of this project is to help you gain experience with the collection of data. The presentation of the data, and the analysis of the data. We also needed to choose appropriate techniques for answering questions. We applied all the statistical tools in our project studied during the whole session. The main basic objective of doing this project was to understand the use of the statistical tools in the real world.
Page 4 of 78 FINAL REPORT OF BQT
Table of Contents:
Sr No Contents Page no1 Authors Detail 12 Acknowledgement 23 Objective 34 List of Tables 55 List of figures 66 List of Abbreviations 77 History of Jazz 88 History of Ufone 99 History of Telenor 1010 Description of Statical Tools 11-1211 Description of Variables 1312 Descriptive Analysis 18-2913 Analysis of Variance 30-4314 Correlation Analysis with Hypothesis 44-5115 Regression Analysis with Hypothesis 52-6416 Comparison Among Companies 65-6617 Crux of Study 6718 Recommendation and Suggestions 6819 Conclusion 6920 Appendix 7021 Supporting Material 71
Page 5 of 78 FINAL REPORT OF BQT
List of Tables:
Sr no Contents Page no1 Table For Raw Data of Jazz 142 Table For Raw Data of Ufone 153 Table For Raw Data of Telenor 164 Table For Mean of Jazz 185 Table For Mean of Ufone 196 Table For Mean of Telenor 207 Table For Standard Deviation of Jazz 238 Table For Standard Deviation of Ufone 249 Table For Standard Deviation of Telenor 2510 Table of ANOVA For Advertisement 3311 Table of ANOVA For Customers 3712 Table of ANOVA For Sims Sold 4113 Table of Correlation of Jazz 4414 Table of Correlation of Ufone 4515 Table of Correlation of Telenor 4616 Table of Multiple Regression of Ufone 5217 Table for Multiple Regression of ANOVA 5618 Table of Multiple Regression of Jazz 5719 Table for Multiple Regression of ANOVA 6020 Table of Multiple Regression of Telenor 6121 Table for Multiple Regression of ANOVA 64
Page 6 of 78 FINAL REPORT OF BQT
List of Figures:
No Figures
Page 7 of 78 FINAL REPORT OF BQT
List of Abbreviations:
X1 = Advertisement (Independent Variable)X2 = no of Customers (Independent Variable)y = Sims Sold in units (Dependent Variable)
= is the arithmetic mean of the values.
S.D = Standard deviation
M.D = Mean Deviation
C.V = coefficient of variance
Sk = Skewness
µ
1
= mean of company one
µ
2
= mean of second company
µ
3 =
mean of third company
SST = Total Sum of square
SSR = Regression Sum of square
SSE = Error Sum of square
k= no of categories
Df = Degree of freedom
n= is the no of observation in the sample.
Page 8 of 78 FINAL REPORT OF BQT
= indicates the absolute value.
a = the y- intercept of the estimated plane which is the best point estimate of α.
b 1 = The slope of the estimated plane with x1 which is the best point estimate of β1 and it is changed by the value of x1.
b 2 = The slope of the estimated plane with x2 which is the best point estimate of β2 and it is changed by the value of x2.
HISTORY OF JAZZ
Pakistan Mobile Communications Limited, better known as Mobil ink GSM, is a telecommunication service provider in Pakistan. The company is Pakistan's leading cellular operator with a subscriber base of 31.5m and market share of 31% in October 2010.
Mobil ink’s Head office is located at Mobil ink House, 1-A Kohistan Road, F-8 Markaz Islamabad. Mobil ink’s corporate postpaid package is sold under the brand name "Indigo" and prepaid by the name of "Jazz".
Mobil ink started operations in 1994 as the first GSM cellular Mobile service in Pakistan by MOTOROLA later it was sold to Orascom, an Egypt-based multi-national company.
In addition to cellular service, the Orascom group is diversifying its service portfolio by setting up new businesses and also expanding through acquisitions. Recently, they started offering DSL broadband through a wholly owned subsidiary, Link.Net.
In addition to Mobil ink, the Orascom group also owns TWA (Trans World Associates) which operates an undersea fiber optic cable from Karachi to Fujairah, UAE.
Till June'10 the company had issued two listed bonds to the tune of Rs. 3.2 billion and Rs. 6 billion
In 1990, Pakistan introduced its first mobile phone service called “Paktel”. After eleven years in January 2001, a new cellular company with GSM technology came into the market and they named it Ufone. The government of Pakistan granted them the license of Pak Telecom mobile limited to operate GSM 9000 all Around Pakistan. Ufone succeeded to establish itself in the market by providing quality service at low rates. After it’s opening, initially they started their service in major cities like Rawalpindi, Islamabad, Karachi and Lahore and on major highways. Later the service expanded to other major cities like Peshawar, Quetta and different towns. Ufone invested around $ 65 million to setup its modern technical infrastructure to provide high quality crystal clear voice and fast data transmission. Ufone targeted the middle class, by introducing low rates and different packages with Urdu names like Jazba etc to attracted lower and middle class people. Ufone has expanded its customer support in a very organized manner with a planned network of dealers, outlets for people convenient. Ufone is committed to care for its customers even after they have acquired a new connection. Ufone not only focused on the technical capabilities, but to provide subscribers a convenient to get connections and services. Ufone recently started GPRS through which users can connect themselves to the global village via Internet, also introduced multimedia messaging services and worldwide SMS at flat rates. Not only SMS but also provides Ufone Info service through which user can get latest information about news, sports, religion, horoscope, stock exchange etc even live sports updates. The tariff packages have been designed keeping in mind the requirements of every segment of the society may that be a housewife, a taxi driver, a trader or a student. Ufone started its operation from Islamabad on 29th January 2001. Cellular industry performance in Pakistan before the launch of Ufone was quite dismal, with one of the lowest population penetration rates in the Whole region. Ufone’s strategy from the day one was to change this scenario and ensure that mobile phones are turned into an everyday business and personal communication tools for all.
Page 10 of 78 FINAL REPORT OF BQT
History of Telenor:
Telenor Pakistan is 100% owned by the Telenor Group, an international provider of high quality voice, data, content and communication services in 14 markets across Europe and Asia. Telenor Group is among the largest mobile operators in the world with over 195 million mobile subscriptions (Q3 2010) and a workforce of approximately 34,000.
Telenor Pakistan is the country's single largest European investor, with investments in excess of US$2 billion. It acquired a GSM license in 2004 and began commercial operations on March 15, 2005.
At the end of October 2010 it had a reported subscriber base of 24.12 million, and a market share of 24% making it the country's second largest mobile operator.
Page 11 of 78 FINAL REPORT OF BQT
.Brief description of all Statical Tools:
We apply Different Statical tools on our Project e.g.1: Mean 2: Median 3: Standard deviation 4: Mean deviation 5: coefficient of variance6: skewness7: Regression Analysis8: ANOVA
Mean shows the average means it shows the average sale per month, the mean tells us the company’s average per month and it helps in comparing one company to the other.While median also tells us the average but the difference between the mean and the median is Mean is appropriate when data is symmetrical, because some month’s has less sale or some has greater so for that condition mean is not the appropriate tool then we use median of data having outlay in it because an outlay leads the data at extreme. Thus to overcome this problem we use median to calculate average.It is also the type of average but we calculate it when the data is non- symmetrical.Standard Deviation tells that how much the data is scattered about the mean, the smaller S.D located close to the mean conversely a large S.D reveals that the observations are widely scattered about the mean.The M.D shows how much the data deviates from its mean.When we find that what percent deviation is from mean? Then we use the rule to measure this coefficient of variation that tells how much percent the data deviates from the mean.Skewness tells us that whether the data is Right Skewed or the left skewed. Skewness > 0 - Right skewed distribution - most values are concentrated on left of the
mean, with extreme values to the right. Skewness < 0 - Left skewed distribution - most values are concentrated on the right of
the mean, with extreme values to the left. Skewness = 0 - mean = median, the distribution is symmetrical around the mean.
Page 12 of 78 FINAL REPORT OF BQT
Anova is a procedure use to test the null hypothesis that the mean of three or more populations are all equal. The procedure we are discussing is called the analysis of variance because the test is based on the analysis of variance in the data obtained from different samples, the application of the one way ANOVA requires the following assumptions, hold true
ASSUMPTIONS OF ONE WAY ANOVA
The population from which the samples are drawn are (approximately) normally distributed
The population from which the samples are drawn have the same variance or standard deviation
The samples drawn from different populations are random and independent.
The ANOVA test is applied by calculating two estimates of the variance of the population distribution the variance between samples and the variance within samples. The variance between samples is also called mean square between sample or MSB. The variance within samples is also called the mean square within sample or MSW.
The variance between samples gives an estimation of variance based on the variation among the means of the samples taken from the different population.
If the means of all population under consideration are equal the means of the respective sample will still be different but the variation among them is expected to be small and consequently the value of MSB is expected to be small. However the mean of the population under consideration are not equal then the value of the MSB is expected to be large. The one way ANOVA test is always right tail with the rejection region in the right of the F distribution curve.
And the Last one is the Multiple Regression and the Correlation that tells us about the Relationship between the dependent and the independent variable and the value of correlation tells us whether the relationship between the dependent and the independent variable is Strong positive or negative and how much the model is best fitted.
Page 13 of 78 FINAL REPORT OF BQT
Brief description of Variables:
We have selected three Companies Ufone, Jazz and Telenor and also Variables that are
1. Advertisement2. No of Customers3. No of Sims Sold
Independent:
Advertisement:
Advertisement is our independent variable and it effects on our sale. No of customers:
No of customers is an independent variable and it also effects on our sale.
Dependent:
No of Sims Sold:No of Sims sold is our dependent variable and it depends upon Advertisement and no of customers.
We collected the thirty months data of these variables from the sales service center of
Ufone, Jazz and Telenor, So that we can apply all the Statical tools on the data through
these three variables.
Page 14 of 78 FINAL REPORT OF BQT
As the Advertisement and no of customers are the independent variable and Sims Sold
are the dependent variable so to check the Relationship between these variables we apply
the Regression and the correlation Analysis on the thirty months data to check whether
the Sims Sold and the Advertisement, no of customers have Direct Relation or the in
direct Relation means for example Weather the sale will increase with the increase in the
Advertisement or not?Raw Data of jazz
Sr No
Year Month Advertisement No. Of Customers No. of Sims Sold
1 2008 June3751000 15339 85551
2 2008 July 3432000 14000 813303 2008 Aug 3000000 14530 785554 2008 Sept 2721000 13880 653305 2008 Oct 3545000 16580 771006 2008 Nov 3233000 14775 733807 2008 Dec 2843000 12335 613318 2009 Jan 2900000 13557 645239 2009 Feb 3100000 14330 6738210 2009 Mar 3300000 15000 6967011 2009 April 3200000 15600 7000012 2009 May 3500000 15890 7188513 2009 June 3700000 16338 7232014 2009 July 3951000 17870 7412315 2009 Aug 3104000 13280 7030016 2009 Sept 3211000 14000 7155017 2009 Oct 3623000 15330 73115
Page 15 of 78 FINAL REPORT OF BQT
18 2009 Nov 3737000 16110 7400019 2009 Dec 3508000 15870 7300020 2010 Jan 2915000 13320 7110821 2010 Feb 2826000 12890 6980022 2010 Mar 2918000 13115 7000023 2010 April 3000000 14310 6890024 2010 May 3371000 14990 8600025 2010 June 3547000 15000 8755026 2010 July 3725000 15670 9000027 2010 Aug 3111000 13800 8050028 2010 Sept 2715000 12000 7589029 2010 Oct 2803000 13200 7870030 2010 Nov 3408000 14500 85370
Raw Data of UfoneSr No Year Month Advertisement No. Of Customers No. of Sims Sold
1 2008 June 3832000 6750 440002 2008 July 3730000 6160 42550
3 2008 Aug 3500000 7120 41800
4 2008 Sept 3808000 5110 39310
5 2008 Oct 3929000 4220 43880
6 2008 Nov 3708000 4000 40000
7 2008 Dec 3800000 5300 43800
8 2009 Jan 3513000 5050 40000
9 2009 Feb 3617000 5711 42900
10 2009 Mar 3615000 6150 41670
11 2009 April 3822000 7033 42100
12 2009 May 4800000 7881 49750
13 2009 June 4908000 8100 52810
14 2009 July 4970000 7750 40760
15 2009 Aug 3800000 6952 43752
16 2009 Sept 3600000 6509 40000
17 2009 Oct 3700000 6800 40710
18 2009 Nov 4000000 6000 46000
19 2009 Dec 4325000 7220 48751
Page 16 of 78 FINAL REPORT OF BQT
20 2010 Jan 4400000 6315 35000
21 2010 Feb 4500000 6714 37000
22 2010 Mar 4313000 5833 39980
23 2010 April 4633000 6124 45609
24 2010 May 4750000 6800 48000
25 2010 June 4831000 7810 50000
26 2010 July 5000000 7650 42660
27 2010 Aug 4625000 4553 40000
28 2010 Sept 4700000 5720 41678
29 2010 Oct 4653000 5209 44749
30 2010 Nov 4900000 3880 50909
Raw Data of TelenorSr No
Year Month Advertisement No. Of Customers No. of Sims Sold
1 2008 June 2700000 8980 920002 2008 July 2800000 9000 935003 2008 Aug 1500000 8500 800004 2008 Sept 2000000 8750 810005 2008 Oct 2500000 7000 830006 2008 Nov 2508850 6500 700007 2008 Dec 2500000 6300 670008 2009 Jan 2900000 5500 650009 2009 Feb 2100000 5800 4500010 2009 Mar 2000000 5300 4185011 2009 April 2200000 5600 4200012 2009 May 2800000 5750 4375013 2009 June 2600000 5980 4000014 2009 July 2709800 6000 4100015 2009 Aug 2000000 5600 4100016 2009 Sept 2100000 6000 4389017 2009 Oct 2800000 6400 4300018 2009 Nov 2600000 6100 4100019 2009 Dec 2200000 6050 4077020 2010 Jan 2400000 6090 4165021 2010 Feb 1900000 4000 3600022 2010 Mar 2100000 4200 3800023 2010 April 1800000 4500 3300024 2010 May 2000000 4600 34000
Page 17 of 78 FINAL REPORT OF BQT
25 2010 June 2400000 4800 3500026 2010 July 2000000 4500 3500027 2010 Aug 2400000 5000 3800028 2010 Sept 2400000 5100 4000029 2010 Oct 2700000 5400 5400030 2010 Nov 2500000 5300 50000
Explanation: of Raw Data:We collected this data from the head offices of Jazz, Ufone, Telenor in the Gujranwala Region of last 30 months of three variables that we have selected for our project, In these variables Advertisement and no of customers are independent while the no of Sims sold per month is dependent variable, then we divide the data of Advertisement and the no of Sims sold by 1000 for making our calculation easy then we apply the all the business quantitative techniques on this data.
Mean shows the averagely advertisement of Jazz in Gujranwala region.Mean of advertisement is 3257000 in actual because we divide the data by 1000.so the Mean of advertisement of jazz is 3257 that shows the average advertisement per month in Gujranwala regionAnd the mean of no of customers is 14.58 that show averagely 14580 persons visit the sales service center of jazz in a month.Mean of Sims sold is 74.60 this also divided by 1000 so the mean is 74600 that shows jazz sales averagely 74600 Sims per month.
Comparison:
Page 23 of 78 FINAL REPORT OF BQT
Averagely advertisement of these three companies jazz, ufone and telenor are 3257000, 4210000 and 2337000 respectively. Ufone spend more money on advertisement than the jazz and telenor.The strength of the no of customers of jazz that is 15000 is greater than Ufone and telenor.Averagely no of Sims sold of jazz, ufone and telenor are 75000, 43000 and 51000 respectively. sales of jazz is comparatively high because most of the people like business mans and households use the jazz network and also because of this reason that jazz is first telecommunication company in Pakistan. Secondly the telenor average sales are 51000 and greater than the ufone’s sales because they sale directly from their sales service center in Gujranwala.
Median Formula Median = (n+1) th value
Page 24 of 78 FINAL REPORT OF BQT
2
Variables Advertisement No of Customers Sims SoldJazz 3222 14.5 73.05
Ufone 4156.5 6.2375 42.605
Telenor 2400 5.775 41.925
Interpretation:
Mean is appropriate when data is symmetrical, because some month’s has less no of unit’s sold or some has greater so for that condition mean is not the appropriate tool then we use median of data having outlay in it because an outlay leads the data at extreme. Thus to overcome this problem we use median to calculate average.It is also the type of average but we calculate it when the data is non- symmetrical.
Comparison:
Averagely advertisement of these three companies jazz, Ufone and telenor are 3222000, 4156500 and 2400000 respectively. Ufone spend more money on advertisement than the jazz and telenor.The strength of the no of customers of jazz that is 15000 is greater than Ufone and telenor.Averagely no of Sims sold of jazz, Ufone and telenor are 73000, 43000 and 41000 respectively. sales of jazz is comparatively high because most of the people like business mans and households use the jazz network and also because of this reason that jazz is first telecommunication company in Pakistan. There was large out layer in the data of telenor and when we apply median on the data than the average Sims sold of telenor is 4190 which is comparatively less than the sales of Ufone.
Standard Deviation tells that how much the data is scattered about the mean, the smaller S.D located close to the mean conversely a large S.D reveals that the observations are widely scattered about the mean.S.D of Advertisement of Ufone is 511000 and its mean is 4210000 that shows they can advertise 511000 more than its mean or less.We divide our data by 1000 that’s why in actual scenario company can advertise 345000 more or less from its mean that are in actual is 42, 10,000.S.D of no of customers is 1150 and the mean is 6214 that shows 1150 customers more or less can visit the sales service center from the mean.And the S.D of Sims sold are 4.5 and mean is 43 it was also divided by 1000 so company can sale 4500 Sims more or less from the mean that is in actual 43000.
Comparison:
Advertisements of Ufone is more deviated from its mean than jazz and telenor because Ufone focus and spend more on advertisement and no of customer of ufone are less deviated from its mean and sales of telenor are more deviated from its mean value.
Coefficient of Variance
Page 29 of 78 FINAL REPORT OF BQT
Formula
Variables Advertisement No of Customers Sims SoldJazz 11% 9% 9%Ufone 12% 18% 10%Telenor 15% 22% 37%
Interpretation:
When we want to find that how much deviation is from mean in percentage, then we use coefficient of variation to measure this.
E.g. the C.V of Advertisement of jazz is 11% that tells us 11% deviation is from mean because mean does not give us the accurate value so we use it. And c.v of no of customers and Sims sold are 9% and 9% respectively that tells 9%data of both deviates from their means.
Comparison:
Advertisement of telenor is more deviated in percentage from its mean than jazz and Ufone and no of customer of jazz are less deviated from its mean and sales of telenor are more deviated from its mean value. this is because of the difference between the means of these three companies.
Page 30 of 78 FINAL REPORT OF BQT
Skewness Formula
Variables Advertisement No of Customers Sims Sold
Jazz 0.30 0.18 0.66Ufone 0.32 -0.06 0.27
Telenor -0.55 0.41 1.42
Interpretation:
The Skewness of advertisement, no of customers and no of Sims sold of jazz 0.30, 0.18 and 0.66 that is greater than zero and it is Right skewed - most values are concentrated on left of the mean, with extreme values to the right.
Comparison:
Skewness values of Advertisement of jazz and Ufone shows that both are right skewed and advertisement of telenor is left skewed. no of customers of jazz and telenor is right skewed and no of customers of Ufone is left skewed. Sales of these three companies area right skewed.
Page 31 of 78 FINAL REPORT OF BQT
Analysis of Variance:
Scenario:As the Analysis of the variance procedure is used to test the null hypothesis that the mean of the three or more population are the same against the alternative hypothesis that not all population mean are same. So here we want to test that the mean of the advertisement of the three companies Telenor jazz and Ufone are same or not?
ANOVA for advertisement
Step 1:- State the null and the alternative hypothesis
H
O
:
µ
1
= µ
2
= µ
3
H
1:
At least mean of one of the company’s Advertisement is not same.
Step 2:- Select the distribution to use
Because we are comparing the means for three normally distributed populations; we use
the F distribution to make this test.
Step 3:- Determine the rejection and the non rejection region
If f
cal >
f
tabulated
Then we reject H
O
that means we are in the favor of H
1
that is At least the Sims sold of
one of the company is not same.
Page 32 of 78 FINAL REPORT OF BQT
V
1
(K-1) = 3-1 = 2
V
2
(N-k) = 90-3 = 87
f
tabulated
= v
1
, v
2
, α
= 2, 87, .05
F
tabulated
=
3.101296
Calculating the value of the test statistics:
The value of the test statistics F for a test of hypothesis using ANOVA is given by ratio of two variances’ the variance between sample and the variance within sample
Test statistics f for a one way ANNOVA:
Page 33 of 78 FINAL REPORT OF BQT
The value of test statistics F for ANNOVA test is calculated as fallows
F=variance between sample/variance within sample =MSB/MSW
Source of Variance
df SS MS F(calculated)
Between Sample k-1 SSB MSB=SSB/k-1f=MSB/MSWWithin sample n-k SSW MSW=SSW/n-k
Total n-1 SST MST=SST/n-1
In ANNOVA terminology the three method use to find arithmetic means are called treatments the table contains the data on the no of customers of three companies.
Analysis of the variance procedure is used to test the null hypothesis that the mean of three or more populations are the same against the alternative hypothesis that not all populations mean are the analysis of variance can be used to compare two populations mean. In our project case we want to know that the means of the three companies are the same or not. First we want to check the mean of the no of customers of three companies, Telenor, Ufone, Jazz, To test these companies no of customers mean we test the null hypothesis That is H0: µ1 =µ2= µ3(all three population mean are equal)Against the alternative hypothesis H1: At least one population mean are not equalFor this test we are using ANOVA, short for the analysis of variance provides such a procedure. It is used to compare three or more population in a single test.X=no of customers of company K=no of different samples n1= size of the sample one n= number of values in all the samples =n1+n2+n3+……….∑x = the sum of the values in all samples = T1+T2+T3………∑x2 = the sum of the squares of the value of all the samplesTo calculate MSB and MSW we first compute the between-samples sum of squares denoted by SSB. And within sample sum of squares, denoted by SSW. The sum of SSB and SSW is called the total sum of squares and is denoted by SST, that is SST = SSB + SSW.
T1 is obtained by adding the values of the advertisement of the one sample, T2 is obtained by adding the values of the advertisement of the second sample, and we got T3 by adding the values of the advertisement of the third sample.
Page 34 of 78 FINAL REPORT OF BQT
As there are 30 observations in each sample so, n1=n2=n3
Between- and Within-samples sum of squares. The between-samples sum of squares denoted by SSB, is calculated as
Source of Variance
df SS MS F(calculated)
Between Sample 2 52576987.6 26,288,493.8152Within sample 87 15000134.7 172,415.3414
Total 89 67577122.1 759,293.5067
Now, f
tabulated
is 3.1011296 and f
cal is
152 so we reject Ho, we are in the favor of H
1
that
means the mean of advertisement of at least one of the company is not same. In other
words, at least one of the mean is different from the other two means then to find in
which company the mean are different we apply the Fisher LSD method.
Page 35 of 78 FINAL REPORT OF BQT
Fisher LSD
Page 36 of 78 FINAL REPORT OF BQT
So here we use the Fisher LSD method to find that which mean is different from others.
If
> LSD (Then we reject Ho otherwise we don’t reject)
H
0
:µ
1
=µ
2
H
1
: µ
1
≠µ
2
H
0
:µ
2
=µ
3
H
1
: µ
2
≠µ
3
H
0
:µ
1
=µ
3
H
1
:µ
1
≠µ
3
If
>LSD we will reject H0.
= 1.96 (106.67) = 209.07
952.8>209.07So, We Reject Ho, we are in the favor of H1.
1872.1>209.07
Page 37 of 78 FINAL REPORT OF BQT
So, We Reject Ho, we are in the favor of H1.
919.3>157.60So, We Reject Ho, we are in the favor of H1.
Conclusion: It means we are in the favor of H1, that means mean of Advertisement of all the three companies are different so we reject Ho.
Page 38 of 78 FINAL REPORT OF BQT
Scenario:As the Analysis of the variance procedure is used to test the null hypothesis that the mean of the three or more population are the same against the alternative hypothesis that not all population mean are same. So here we want to test that the mean of the no of customers of the three companies Telenor jazz and Ufone are same or not?
ANOVA for the no of customers
H
O
:
µ
1
= µ
2
= µ
3
H
1:
At least mean of one of the company’s no of customers is not same.
If f
cal >
f
tabulated
Then we reject H
O
that means we are in the favor of H
1
that is, at least the no of
customers of one of the company is not same.
f
t ab=
(v1
,
v2, α)
=2, 87, 0.05
= 3.101296
If f calculated > 3.101296 we will reject H
O.
Page 39 of 78 FINAL REPORT OF BQT
Source of variation df SS MS fBetween sample 2 1444853152
722426576
43.5268Within sample 87 144396252.4
16597270.39Total89
15892494
178,567.348
If f calculated > 3.101296 we will reject H
O,
we are in the
favor of H
1
means at least mean of one of the company’s no of customers mean is
different from the other two means then to find in which company the mean are different
we apply the Fisher LSD method.
Page 40 of 78 FINAL REPORT OF BQT
Fisher LSD
If
>LSD we will reject H0.
Page 41 of 78 FINAL REPORT OF BQT
1.96 110537.8
1.96(33.4)
651.64
=8366.16667 and LSD=651.64
We will reject H0 because >LSD
=260.8
We will not reject H0 because <LSD
=8626.966
We will reject H0 because >LSD
Page 42 of 78 FINAL REPORT OF BQT
Conclusion: It means we are in the favor of H1 that at least one companies’ mean of no of customers is not same means we reject HO all the three means of no of customers are different.
Scenario:As the Analysis of the variance procedure is used to test the null hypothesis that the mean of the three or more population are the same against the alternative hypothesis that not all population mean are same. So here we want to test that the mean of the no of Sims sold of the three companies Telenor jazz and Ufone are same or not?ANOVA for Sims sold
H
O
:
µ
1
= µ
2
= µ
3
Page 43 of 78 FINAL REPORT OF BQT
H
1:
At least mean of one of the company’s no of customers is not same.
If f
cal >
f
tabulated
Then we reject H
O
that means we are in the favor of H
1
that is, at least the no of
customers of one of the company is not same.
f
t ab=
(v
1,
v
2
, α)
=2, 87, 0.05
= 3.101296
If f calculated > 3.101296 we will reject H
O.
Page 44 of 78 FINAL REPORT OF BQT
Calculated Value:Source of variation df SS MS f
Between sample 2 15946.0 797358.495Within sample 87 11858.5 136.30
Total 89 27804.5 312.41
So, fcal =58.495> f tabulated=3.101296 therefore we reject H
o
we are in the favor of H
1
that is At least average of one of the company of Sims sold is different from the other two
means then to find in which company the mean are different we apply the Fisher LSD
method.
Page 45 of 78 FINAL REPORT OF BQT
Fisher LSD
If
> LSD
Then we reject Ho otherwise we don’t reject.
H0:µ
1
=µ2 H1: µ
1
≠µ
2
H0:µ
2
=µ
3
H1: µ
2
≠µ
3
H0:µ
1
=µ
3
H1:µ
1
≠µ
3
If
Page 46 of 78 FINAL REPORT OF BQT
= 5.8604
Page 47 of 78 FINAL REPORT OF BQT
>LSD we will reject H0.
Page 48 of 78 FINAL REPORT OF BQT
Page 49 of 78 FINAL REPORT OF BQT
= and LSD=5.8604
We will reject H0 because >LSD
=-7.64
We will not reject H0 because <LSD
=23.59666
We will reject H0 because >LSD
Conclusion : It means we are in the favor of H1 and we reject HO that means all the three means of no of Sims sold are different.
The “r” tells that which relation occurs between the two variables. Our calculated value of the “r” is showing that there is a strong positive relationship between Advertisement, No of customers and the Sims sold.R is the measure of correlation between the observed value and the predicted value of the criterion variable. In our project this would be the correlation between the no of Sims sold actual and no of Sims sold predicted by our predicted variables. R2: This is the measure of how good a prediction of the criterion variable we can make by knowing the predictor variable.r2 is the value of explained variation.What we forecasted through regression equation and find the error between actual value and predicted value we can reduce that error by explained variation.R2 tells us that how much percent our model is best fitted.SSR = Explained VariationSSE= Unexplained VariationBasically r2 tells us the accuracy of our model or how much the model is best fitted from the regression equation.
Page 54 of 78 FINAL REPORT OF BQT
Ufone:
In our scenario of Ufone the value of r2 is 21% it means that Regression Equation can minimize 21% of the error or it has 21% explained variation, remaining 79% are the other factors are also included that are unexplained variation.
Adjusted r2
is the improved version of the r2
and it is over estimate the value so we find r2
which tells the error in Reality and it is always lesser than the r2
.
So, in our scenario value of adjusted r2
is 12% that means 12% error in the Regression
Equation in Reality.
Jazz:
In our scenario of Jazz the value of r2 is 35% it means that Regression Equation can minimize 35% of the error or it has 35% explained variation, remaining 65% are the other factors are also included that are unexplained variation.
So, in our scenario value of adjusted r2
is 27% that means 27% error in the Regression
Equation in Reality.
Telenor:
In our scenario of Telenor the value of r2 is 75% it means that Regression Equation can minimize 75% of the error or it has 25% explained variation, remaining 25% are the other factors are also included that are unexplained variation.
Page 55 of 78 FINAL REPORT OF BQT
So, in our scenario value of adjusted r2
is 74% that means 74% error in the Regression
Equation in Reality.
Hypothesis of correlation of Ufone :
We want to check the relation between the Advertisement and the no of customers….
T-distribution
We reject h0 if t>2.05 and t<-2.05
We reject h0 because t>2.05 we are in the favor of H1 that means there is the Linear Relationship between the Dependent and the independent variable.
Page 56 of 78 FINAL REPORT OF BQT
If Advertisement and the no of customers increase the no of Sims Sold will also increase.
Hypothesis of correlation of Jazz:
If there is a significant linear relationship between the independent variable and the
dependent variable, the value of will not equal zero.
T-distribution
We reject h0 if f>2.05 and f<-2.05
Page 57 of 78 FINAL REPORT OF BQT
We reject h0 because f>2.05 we are in the favor of H1, so there is the Linear Relationship between the Dependent and the independent variable.If Advertisement and the no of customers increase the no of Sims Sold will also increase.
Hypothesis of correlation of Telenor:
If there is a significant linear relationship between the independent variable and the
dependent variable, the value of will not equal zero.
T-distribution
We reject h0 if t cal >2.05 and t<-2.05
Page 58 of 78 FINAL REPORT OF BQT
. We reject h0 because t >2.05 we are in the favor of H1, so there is the Linear Relationship between the Dependent and the independent variable.If Advertisement and the no of customers increase the no of Sims Sold will also increase.
is called the Multiple Regression equation.Y is the dependent variable or the predicted variable.
And x1 and x2 is the independent variable or it is called the predicted variable.X1= Advertisement (Independent Variable)X2= No of customers (Independent Variable)Y= Sims sold (Dependent Variable)
is the projected or forecasted value.We find the value of “e” because this value tells us the difference between the actual and the projected value.“a” is the fixed value and “b” is the per unit change, a = the y- intercept of the estimated plane which is the best point estimate of α.B 1 = the slope of the estimated plain with x1 which is the best point estimate of β1 and it is changed by the value of x1.B 2 = the slope of the estimated plain with x2 which is the best point estimate of β2 and it is changed by the value of x2.
The value of “a” tells us that if the x1 and x2 that is Advertisement and no of customers is zero then how much it is affected on the sale and it is fixed or constant so there is no change in the value of a when x1 or x2 is increase or decrease.B is the per unit change which change with the value of x1 and x2 that are our independent variables.
Now the regression equation is:Y = a + b1x1 + b2x2 + b3x3
In this equation Advertisement and the no of customers have the positive relation with no of Sims sold that means if Advertisement and no of customers will increase then no of Sims sold will also increase.
Page 61 of 78 FINAL REPORT OF BQT
Hypothesis Testing of Multiple Regression of Ufone:
The test is used to check whether if there is a significant linear relationship between the
independent variable and the dependent variable, if yes then the slope will not equal zero.
HO: β0= 0H1: β0≠ 0
B1 = 0.003094
Page 62 of 78 FINAL REPORT OF BQT
t tab. v, α = 27, 0.05 = 1.703288
We Reject H0 if t > 1.70.
T cal = 4.827
T > 1.70 so, we Reject H0, we are in the favor of H1 it means the value of b1 is not equal to the zero and there is the Perfect Relationship between the Advertisement and the no of Sims Sold.
B 2 = 0.536377
T tab. v, α = 27, 0.05 = 1.703288We Reject H0 if t > 1.70.
T cal = 1.243
T Cal< 1.70 so, we do not Reject H0, we are in the favor of H0 means the value of b2 is very nearly equal to the zero and there is no relationship between the No of customers and the no of Sims sold.
Multiple regressions ANOVA testing of Ufone:ScenarioAs the Analysis of the variance procedure is used to test the null hypothesis we want to check that the Advertisement and No of customers have direct relation or not means with the increase in the Advertisement no of customers will increase or not, or how much per unit change will occur on b1 with the change in b2, B1 Represents the Advertisement and B2 Represents the No of customers,Step1:
H0: b1 = b2
H1: b1 ≠ b2 Step2: F-Distribution
Page 63 of 78 FINAL REPORT OF BQT
Step3:V1 = 3V2 = 26
If f cal is > 2.9752 then we reject H0 Otherwise we do not reject H0.
Step4:
Source of variation
Degree of freedom
Sum of squares (SS) Mean squares (MS)
Value of test statistics (f)
regression K -1= 2 SSR= 106 MSR=SSR/k = 53
F=MSR/MSE = 3.5584
Error n-k30- 3= 27
SSE= 402.1382 MSE=SSE/n-(k+1)= 14.894
Total 29 SST =508.36 MST=SST/n-1= 18
Step5:
Decision rule:
F cal is > 2.9752 so; we reject H0 we are in the favor of H1 that means with the increase in the Advertisement, no of customers do not increase.
In this equation Advertisement and the Sale has the positive relation and N o of customers and Sims sold has the negative relation that means If Advertisement will increase then no of Sims sold will also increase.
Hypothesis Testing of Multiple Regression of Jazz:
HO: β0= 0H1: β0≠ 0
B 1 = 0.02178
t tab. v, α = 27, 0.05 = 1.703288We Reject H0 if t > 1.70.
T cal = 14.449
T > 1.70 so, we Reject H0, we are in the favor of H1 that means Advertisement and no of Sims Sold have linear Relationship.
HO: β0= 0H1: β0≠ 0
B 2 = -3.4993
T cal = -48.3997
Page 66 of 78 FINAL REPORT OF BQT
t > 1.70 So, we Do not Reject H0, we are in the favor of H0 that means No of customers and no of Sims sold have no relation.
Multiple regressions ANOVA testing of Jazz:ScenarioAs the Analysis of the variance procedure is used to test the null hypothesis that with the increase in Advertisement, no of customers will increase or not or how much per unit change will occur on b1 with the change in b2, B1 Represents the Advertisement and B2
Represents the No of customers,
Step1:H0: b1 = b2
H1: b1 ≠ b2 Step2:
Page 67 of 78 FINAL REPORT OF BQT
F-DistributionStep3: V1 = 3
V2 = 26
If f cal is > 2.9752 then we reject H0 Otherwise we do not reject H0.
Step4:
Source of variation
Degree of freedom
Sum of squares (SS) Mean squares (MS)
Value of test statistics (f)
Regression K -1 = 2 SSR= 508 MSR=SSR/k = 254
F=MSR/MSE = 7.0006
Error n-k30-3= 27
SSE= 979 MSE=SSE/n-(k+1)= 36.25
Total 29 SST =1501 MST=SST/n-1= 51.75
Step5:
Decision:
f cal is > 2.9752 So, we reject H0, so we are in the favor of H1 so that Advertisement and no of customers have no relation to each other If we increase the Advertisement the no of customer will not increase
In this equation Advertisement and the no of customers have the positive relation with no of Sims sold that means if Advertisement and no of customers will increase then no of Sims sold will also increase.
Hypothesis Testing of Multiple Regression of Telenor:
HO: β0= 0H1: β0≠ 0
B 1 = 0.0107
t tab. v, α = 27, 0.05 = 1.703288We Reject H0 if t > 1.70.
T cal = 14.069
T Cal > 1.70 so, we Reject H0, we are in the favor of H1 that means Advertisement and no of Sims Sold have linear Relationship.
HO: β0= 0H1: β0≠ 0
B 2 = 11.814
t tab. v, α = 27, 0.05 = 1.703288We Reject H0 if t > 1.70.
Page 70 of 78 FINAL REPORT OF BQT
T cal = 40.320t Cal > 1.70 So, we Reject H0, we are in the favor of H1 that means no of customers and no of Sims Sold have linear Relationship.
Multiple regressions ANOVA testing of Telenor:Scenario
As the Analysis of the variance procedure is used to test the null hypothesis we want to check that the Advertisement and No of customers have direct relation or not means with the increase in the Advertisement no of customers will increase or not, or how much per unit change will occur on b1 with the change in b2, B1 Represents the Advertisement and B2 Represents the No of customers,
Step1:H0: B1 = B2
H1: B1 ≠ B2
Page 71 of 78 FINAL REPORT OF BQT
Step2:F-Distribution
Step3:V1 = 3V2 = 26
If f cal is > 2.9752 then we reject H0 Otherwise we do not reject H0.
Step4:
Source of variation
Degree of freedom
Sum of squares (SS)
Mean squares (MS)
Value of test statistics (f)
regression K -1=2
SSR= 7353 MSR=SSR/k = 3676.5
F=MSR/MSE = 40
error n-k30-3= 27
SSE= 2512 MSE=SSE/n-(k+1)= 93.037
total 29 SST =9865 MST=SST/n-1= 340.17
Step5:Decision:
F cal is > 2.9752 so, we reject H0, and we are in the favor of H1 that is Advertisement and no of customers are not equal or with the increase in the Advertisement, no of customers will not increase.
Comparison among the companies on the basis of Regression and Correlation
UfoneIn this equation Advertisement and the no of customers have the positive relation with no of Sims sold that means if Advertisement and no of customers will increase then no of Sims sold will also increase.
T > 1.70 so, we Reject H0, we are in the favor of H1 it means the value of b1 is not equal to the zero and there is the Perfect Relationship between the Advertisement and the no of Sims Sold.
Page 72 of 78 FINAL REPORT OF BQT
T Cal< 1.70 so, we do not Reject H0, we are in the favor of H0 means the value of b2 is very nearly equal to the zero and there is no relationship between the No of customers and the no of Sims sold.We reject h0 because t>2.05 we are in the favor of H1 that means there is the Linear Relationship between the Dependent and the independent variable.If Advertisement and the no of customers increase the no of Sims Sold will also increase.
JazzIn this equation Advertisement and the Sale has the positive relation and N o of customers and Sims sold has the negative relation that means If Advertisement will increase then no of Sims sold will also increase.
T > 1.70 so, we Reject H0, we are in the favor of H1 it means the value of b1 is not equal to the zero and there is the Perfect Relationship between the Advertisement and the no of Sims Sold.T Cal< 1.70 so, we do not Reject H0, we are in the favor of H0 means the value of b2 is very nearly equal to the zero and there is no relationship between the No of customers and the no of Sims sold.
We reject h0 because f>2.05 we are in the favor of H1, so there is the Linear Relationship between the Dependent and the independent variable.If Advertisement and the no of customers increase the no of Sims Sold will also increase.
TelenorIn this equation Advertisement and the no of customers have the positive relation with no of Sims sold that means if Advertisement and no of customers will increase then no of Sims sold will also increase.
T Cal > 1.70 so, we Reject H0, we are in the favor of H1 that means Advertisement and no of Sims Sold have linear Relationship.
t Cal > 1.70 So, we Reject H0, we are in the favor of H1 that means no of customers and no of Sims Sold have linear Relationship.
Page 73 of 78 FINAL REPORT OF BQT
We reject h0 because t >2.05 we are in the favor of H1, so there is the Linear Relationship between the Dependent and the independent variable.If Advertisement and the no of customers increase the no of Sims Sold will also increase.
In Short
The Regression Equation of Ufone and the Telenor tells us that there is perfect relationship between the dependent and the independent variable but the Regression Equation of Jazz tells us that there is positive Relationship between the Advertisement and the no of Sims Sold but the negative relation between the no of customers and the no of Sims Sold
It shows that the if Advertisement and no of Customers increases then the no of Sims Sold will also increase.
Crux of Study
We selected three companies for BQT project (jazz Ufone Telenor) and the three variables for each (two are the independent variables (Advertisement no of customers) and one is dependent (no of Sims sold). We applied different tools and came in to the following findings;Averagely advertisement of these three companies jazz, ufone and telenor are 3222000, 4156500 and 2400000 respectively. Ufone spend more money on advertisement than the jazz and telenor.The strength of the no of customers of jazz that is 15000 is greater than ufone and telenor.
Page 74 of 78 FINAL REPORT OF BQT
Averagely no of Sims sold of jazz, ufone and telenor are 73000, 43000 and 41000 respectively. Sales of jazz is comparatively high because most of the people like business mans and households use the jazz network and also because of this reason that jazz is first telecommunication company in pakistan.there was large out layer in the data of telenor and when we apply median on the data than the average Sims sold of telenor is 4190 which is comparatively less than the sales of ufone.
Advertisements of ufone is more deviated from its mean than jazz and telenor because ufone focus and spend more on advertisement and no of customer of ufone are less deviated from its mean and sales of telenor are more deviated from its mean value. And coefficient of variance tells us these values in percentage.Skewness values of Advertisement of jazz and ufone shows that both are right skewed and advertisement of telenor is left skewed. No of customers of jazz and telenor is right skewed and no of customers of ufone is left skewed. Sales of these three companies are right skewed.In analysis of variance we checked that the mean of the variables of these companies are same or not (it does not mean that the values of the mean will be same it tells that the variation among the mean’s is less or more). After the Anova testing we found we are in the favor of H1 that the mean of all the companies are not the same.Then we did the correlation analysis with hypothesis testing. In our scenario of Ufone the value of r2 is 21% it means that Regression Equation can minimize 21% of the error.Value of adjusted r2 is 12% that means 12% error in the Regression Equation in Reality. And like this the value of the r2 and adjusted r2 for jazz are 35%, 27% and for telenor are 75%, 74%.
Recommendations and Suggestions
By increasing the advertisement the no of customers increased so we recommend to jazz that they should increase their advertisement so they can increase their no of customers as well as sales. Ufone should manage, develop and maintain productive relationship with internal customers and partners. Because of the Denmark issue the repute of Telenor was badly affected so the company should do their best to catch the customer’s loyalty they should offer attractive
Page 75 of 78 FINAL REPORT OF BQT
packages and make more and attractive advertisement to catch customer’s attractiveness. Telenor Company can also increase their sales by enhancing their supply chain distribution.
Conclusions
The no of customers of the jazz are more than Ufone and Telenor because it was the first telecommunication company of the Pakistan and have the largest market share. Mostly the households and business people are using jazz due to their attractive offers like ladies first.The advertisement of the ufone is more than the jazz and telenor because with the increase of the advertisement they attract more customers as compare to the jazz and the telenor. Ufone is the second largest company in the Pakistan has also shown by the values of its customers and the no of Sims sold in the descriptive analysis.
Page 76 of 78 FINAL REPORT OF BQT
The three companies we selected, from these three companies the telenor is at no. 3rd in every aspect as concluded from the values of Advertisement, no of customers, no of Sims sold.