# HI6007 Statistics And Research Methods For Business Decision Making Assignment

(a). Arrange the data of 20 student’s result

 Student number Results 1 42 2 53 3 54 4 61 5 61 6 61 7 62 8 63 9 64 10 66 11 67 12 67 13 68 14 69 15 71 16 71 17 76 18 78 19 81 20 83 Total 1318

Data has arranged in ascending order for getting correct results.

• Compute Mean, Median and Mode

To calculate mean , following formula will implement:

Σ xi /9Total of marks) =1318;    n (number of students)=20

131820= 65.9  is the  mean value of student’s results.

Median calculation

Median = N2+1=202+1=11th  Item of the above table will be median

Median is 67

• Compute 1st and 3rd Quartile

1st Quartile =  14 N+1 = 20+1/4 = 5th Item  = 61

3rd Quartile =34N+1=3420+1=634=15th  item=71

• Compute and Intercept 90th Percentile

Formula = 90% of total number of students(observation)  =  0.90*20 = 18th Item = 78

(b) Inferential Statistics:

Inferential statistics include select sample from the available observation, in order to identify or solve the  issue through applying appropriate tests. These tests are also helpful in hypothesis testing and prove the outcome with valid numbers.

(i) Prepare Joint Probability Table

 Applied for More than 1 University Age Group Yes No 23 and Under 207 207/808*100=25.62% 201 201/1210*100=16.61% 24-26 299 299/808*100=37.0% 379 379/1210*100=31.32% 27-30 185 185/808*100 = 22.90% 268 268/1210*100=22.15% 31-35 66 66/808*100=8.17% 193 193/1210*100= 15.95% 36 and over 51 51/808*100=6.31% 169 169/1210*100= 13.97% Total / Joint Probability(%) 808 100% 1210 100%

(ii) Given that a student applied to more than 1 university, what is the probability that the student is 24-26 years old.

Probability of student is 24-26 years old = 299/808 =37.00%

• Is the number of universities applied to independent of student age? Explain

Student age  is an independent variable against number of observation collected for application made for enrolment in more than one university at a time. Any student at any age can take enrolment of in more than one university or they can adopt only one university at a time.

(b)

Information provided-

 x f(x) 10 0.05 20 0.1 30 0.1 40 0.2 50 0.35 60 0.2 Total

X represent number of new clients for counselling cases in the year 2021.

Formula of calculating Expected value =  𝐸(𝑥) = 𝜇 = ∑𝑥 𝑓(𝑥)

 x f(x) (𝑥 − 𝜇) 2 (𝑥 − 𝜇) 2*f(x) 10 0.05 -33 1089 54.45 20 0.1 -23 529 52.9 30 0.1 -13 169 16.9 40 0.2 -3 9 1.8 50 0.35 7 49 17.15 60 0.2 17 289 57.8 Total 201

Expected Value= (10*0.05+20*0.1+30*0.1+40*0.2+50*0.35+60*0.2) =43

Formula of Variance of a discrete random variable 𝑉𝑎𝑟(𝑥) = ∑(𝑥𝜇) 2 𝑓(𝑥

Variance = 201 (calculation shows in table)

1. Formulate Hypothesis :

Problem statement: Population annual expenditure on prescription drugs per person is lower in the Midwest than the Northeast.

Hypothesis Statement:

Ho: µ ≤ \$838 or

Ho: µ = \$838

Ha: µ > \$838

Problem statement can test on one tail test from left  tail as it requires testing of lower limit.

• Suitable test Statistics

One (Left) tail test

Formula:

• Calculate value of relevant test statistics and P- value

Sample Mean (x) = \$745

Null Hypothesis Mean  = \$838

SD = 300

Sample size = 60

Applying Formula (745-838)/300/sqrt(n)

Z=  -93/38.75 = -2.40

From the table given of Z score , at significance level of 0.05 ,

P value = 0.0071

• Based on the p value in part (III), at 99% confidence level, decide the decision criteria.

If the confidence level is 99% then there is 1% of significance level  for this problem and at this level  the critical value is  2.326 , for this Z-score is -2.4 which is less than critical value (2.326> -2.4). Null hypothesis shall be rejected.

• Make  the conclusion Based on the analysis.

As per rejection of null hypothesis, it is concluded that the prescribed drugs expenditure is not lower in Midwest as comparison to Northwest.

•  State the null and alternative hypothesis for single factor ANOVA to test for any significant difference in the mean price of gasoline for the three brands.

Hypothesis

H0 = µ1= µ2= µ3

H1 = µ1≠µ2≠ µ3

(ii) State the decision rule at 5% significance level.

Reject the H0 id t stat  > Z critical value, Otherwise accept the null hypothesis

(iii) Calculate the test statistics

 A B C 3.77 3.83 3.78 3.72 3.83 3.87 3.87 3.85 3.89 3.76 3.77 3.79 3.83 3.84 3.87 3.85 3.84 3.87 3.93 4.04 3.99 3.79 3.78 3.79 3.78 3.84 3.79 3.81 3.84 3.86 Sample Mean 3.811 3.846 3.85 Varience 0.003349 0.004844 0.00382

ANOVA one- way test Formula

Formula F= MSTR / MSE

MSTR = 𝑆𝑆𝑇𝑅 / 𝑘 – 1

MSE = SSE /𝑛r – k

𝑥Ӗ= (3.81 + 3.84 + 3.85)/3 = 3.83

SSTR=  10(3.81- 3.83)+ 10(3.84-3.83)2 + 10(3.85-3.83)= 0.009

MSTR = 0.009/ (3-1) = 0.0045

P-value and critical value approaches

Value of test statistic

SSE = 9(0.003) +9(0.005) + 9(0.004) =0.108

MSE = 0.108/(30-3) = 0.004

F= 0.0045/0.004 =1.125

ANOVA Table

 Source of variation Sum of Squares Degrees of Freedom Mean Square F P- value Treatment 0.009 2 0.0045 1.125 0.044 Error 0.108 27 0.004 Total 0.117 29 0.0085

P- value calculation

Here Numerator df = 2; Denominator Df = 27 then the value of F at 0.01 = 5.49

Decision on the basis of test

The p-value < .05, So null hypothesis shall be rejected

Decision as per critical value approach

Based on an F distribution with 2 numerator d.f. and 27 denominator d.f., F.05 = 3.35.

Reject H0 if F < 3.35

Here F = 1.125 <3.35 which is evidence for rejection of null hypothesis.

(d) Based on the calculated test statistics decide whether any significant difference in the mean price of gasoline for three bands.

The value of F is 1.125 which is lower than F critical value  this means  hypothesis has been rejected that means  there is significant difference in the mean price of gasoline in all the three brands.

• Complete the missing entries from A to H in this output

A= R Square = SSR/SST = 35250755.68/ 42699148.82 = 0.82

B= Observation = 50 (provided in Answer)

C= residual = Total- Regression = 49-2 = 47

D= 42699148.82-7448393.14 = 35250755.68

E= SSRegression / dfreg. = 35250755.68/2 = 17625377.8

F= SSR /(50-3) = 7448393.148/47 = 158476.45

G= 17625377.8/158476.45 = 111.217647

H= Coefficient of income /Standard error of income = 8.36

• Estimate the annual credit card charges for a three-person household with an annual income of \$40,000

To estimate  charges of credit card , intercept, household value and size has been considered from the ANOVA table.

The annual credit card charges for three person family is \$3700  where annual income is \$40000.

• Did the estimated regression equation provide a good fit to the data? Explain

No,  The reason behind the same is high variability  between two variables X and Y  which  fails in establishing good fit to the data.

1. Using linear trend equation forecast the sales of face masks for October 2020
 Month Sales (\$) 1 17000 2 18000 3 19500 4 22000 5 21000 6 23000

Linear Trend Equation = Y= Mx+B

M= Y2-Y1X1-X2 = 23000-21000/6-5 = 2000/1 = 2000

Y = Mx+B

Y= ?

X=1

B= 23000

Y = 2000*1+23000

Y = 2000+23000= 25000

\$25000 will be the forecasted sale in the month of October.

• Sales forecast will be
 Sales Weight Weighted Sale July 22,000 0.2 4400 August 21,000 0.3 6300 September 23,000 0.5 11500 Total 22200

So, the expected sale for the next month will be \$22200