Scribd Logo
Upload

Welcome to Scribd!

  • 2K views

    Correlation and Regression Analysis

    Uploaded by

    Manoj Mudiraj
    The document provides data on two studies examining correlations: 1. A study of 8 bank accounts examining the relationship between mean account balance and number of transactions. The correlation coefficient (r) was calculated as 0.0336 and a t-test found insufficient evidence of a correlation greater than zero at the 0.01 significance level. 2. A study of 9 months of advertising expenses and sales revenues for a company. The correlation coefficient (r) was calculated as 0.1971 and a t-test found insufficient evidence of a correlation greater than zero at the 0.10 significance level. 3. Two additional exercises are presented involving correlation analyses but no solutions are shown.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Correlation and Regression Analysis

    Uploaded by

    Manoj Mudiraj
    2K views13 pages
    The document provides data on two studies examining correlations: 1. A study of 8 bank accounts examining the relationship between mean account balance and number of transactions. The correlation coefficient (r) was calculated as 0.0336 and a t-test found insufficient evidence of a correlation greater than zero at the 0.01 significance level. 2. A study of 9 months of advertising expenses and sales revenues for a company. The correlation coefficient (r) was calculated as 0.1971 and a t-test found insufficient evidence of a correlation greater than zero at the 0.10 significance level. 3. Two additional exercises are presented involving correlation analyses but no solutions are shown.

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document provides data on two studies examining correlations: 1. A study of 8 bank accounts examining the relationship between mean account balance and number of transactions. The correlation coefficient (r) was calculated as 0.0336 and a t-test found insufficient evidence of a correlation greater than zero at the 0.01 significance level. 2. A study of 9 months of advertising expenses and sales revenues for a company. The correlation coefficient (r) was calculated as 0.1971 and a t-test found insufficient evidence of a correlation greater than zero at the 0.10 significance level. 3. Two additional exercises are presented involving correlation analyses but no solutions are shown.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    2K views13 pages

    Correlation and Regression Analysis

    Uploaded by

    Manoj Mudiraj
    The document provides data on two studies examining correlations: 1. A study of 8 bank accounts examining the relationship between mean account balance and number of transactions. The correlation coefficient (r) was calculated as 0.0336 and a t-test found insufficient evidence of a correlation greater than zero at the 0.01 significance level. 2. A study of 9 months of advertising expenses and sales revenues for a company. The correlation coefficient (r) was calculated as 0.1971 and a t-test found insufficient evidence of a correlation greater than zero at the 0.10 significance level. 3. Two additional exercises are presented involving correlation analyses but no solutions are shown.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 13

    Name: Vaishnavi peddineni

    Class: BS-Biology New 1-1 (summer)


    Date: 08 July 2021
    Student id:20201513

    Seatwork 6 (Correlation and regression


    Analysis)
    Section Exercise 10.3

    1. AUS Rural Bank is studying the relationship between the mean


    account balance for individual accounts and the number of
    transactions per month. A sample of eight accounts revealed:
    Customer 1 2 3 4 5 6 7 8
    (x)Mean 27 32 50 24 53 10 16 15
    Balance
    (₱000s)
    (y)No. of 11 2 5 9 7 2 4 10
    Transaction
    s
    Find the coefficient of correlation. Determine at the 0.01
    significance level whether the correlation in the population is
    greater than zero.

    Solution:

    Step 1: State the hypotheses.

    𝐻0= The correlation in the population is not greater than zero, 𝜇 ≯0

    𝐻1= The correlation in the population is greater than zero, 𝜇>0

    Step 2: The level of significance and critical region.

    𝜶 =0.01, n=8, df = n-1=8-1=7 and 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = +2.998


    Step 3: Complete the table and compute for the value of r and t.
    Customer x y 𝒙𝟐 𝒚𝟐 xy
    1 27 11 729 121 297
    2 32 2 1024 4 64
    3 50 5 2500 25 250
    4 24 9 576 81 216
    5 53 7 2809 49 371
    6 10 2 100 4 20
    7 16 4 256 16 64
    8 15 10 225 100 150
    Total(n=8) 𝜮(𝒙) = 𝟐𝟐𝟕 𝜮(𝒚)= 50 𝜮𝒙 = 8219
    𝟐
    𝜮𝒚 = 400
    𝟐
    𝜮𝒙𝒚 = 1432
    𝐧𝚺𝐱𝐲−𝚺(𝐱).𝚺(𝐲) r =
    (8)(1432)−(227).(50)

    r= 227) 2][8(400)−(50)2]
    √[8(8219)−

    ( 11456−11350

    r=
    √[65752−51529][3200−2500]
    106

    r=
    √[14196][700] 106

    r=
    √9937200 106
    r=
    3152.3324
    106
    r=
    3152.3324

    r = 0.0336
    = 𝐫×√𝐧−𝟐
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 √𝟏−𝐫𝟐

    tcomputed = √1−(0.0336)0.0336×√8−22
    = √1−0.00112896
    0.0336×√6
    0.0336×2.4494
    tcomputed =
    √0.99887104

    tcomputed tcomputed
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 = 0.0823

    Step 4: Decision rule.

    Reject the null hypothesis, if 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 is greater than the 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍

    Here, 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 < 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 0.0823 < 2.998

    So, we do not have enough evidence against 𝐻0 to reject it, so we

    fail to reject the null hypothesis.

    Step 5: Conclusion.

    At 0.01, significance level we can conclude that the correlation in the


    population is not greater than zero.

    2. SJS Company has been selling to retail customers in the


    Metro Manila area. They advertise extensively on radio, print
    ads, and in the internet. The Owner would to relationship
    between the amount spent on advertising expense and sales
    for the last 9 months.
    Months Jan Feb Mar Apr May June July Aug Sept
    Advertising 10 8 12 11 13 15 14 13 16
    Expense
    Sales 190 215 190 210 235 208 170 175 250
    Reven
    ue
    Find the coefficient of correlation. Determine at the 0.10
    significance level whether the correlation in the population
    is greater than zero.

    Solution:

    Step 1: State the hypotheses.

    𝐻0= The correlation in the population is not greater than zero, 𝜇 ≯0

    𝐻1= The correlation in the population is greater than zero, 𝜇>0

    Step 2: The level of significance and critical region.

    𝜶 =0.10, n=9, df = n-1=9-1=8 and 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = +1.397


    Step 3: Complete the table and compute for the value of r and t.
    Months x y 𝒙𝟐 𝒚𝟐 xy
    Jan 10 190 100 36100 1900
    Feb 8 215 64 46225 1720
    Mar 12 190 144 36100 2280
    Apr 11 210 121 44100 2310
    May 13 235 169 55225 3055
    Jun 15 208 225 43264 3120
    Jul 14 170 196 28900 2380
    Aug 13 175 169 30625 2275
    Sept 16 250 256 62500 4000
    Total(n=9 𝜮(𝒙) = 𝟏𝟏𝟐 𝜮(𝒚)= 1843 𝜮𝒙𝟐 = 1444 𝜮𝒚𝟐 = 383039 𝜮𝒙𝒚 = 23040
    )
    𝐧𝚺𝐱𝐲−𝚺(𝐱).𝚺(𝐲) r =
    (9)(23040)−(112).(1843)

    r= 112) 2][9(383039)−(1843)2]
    √[9(1444)−(
    207360−206416
    r=
    √[12996−12544][3447351−3396649]
    944
    r=
    √[452][50702] 944

    r=
    √22917304
    944 r
    =
    4787.2021

    r = 0.1971

    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 = 𝐫×√𝟏−𝐫
    √𝐧−𝟐
    𝟐

    tcompute =
    d √1−(0.1971)0.1971×√9−
    22
    = √1−0.03884841
    0.1971×√7
    tcompute 0.1971×2.6457
    d
    =
    √0.96115159

    tcomputed tcomputed
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 = 0.5319
    Step 4: Decision rule.

    Reject the null hypothesis, if 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 is greater than the 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍

    Here, 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 < 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 0.5319 < 1.397

    So, we do not have enough evidence against 𝐻0 to reject it, so we

    fail to reject the null hypothesis.

    Step 5: Conclusion.

    At 0.01, significance level we can conclude that the correlation in the


    population is not greater than zero.

    Section Exercise 10.4


    1. The production department of WSS Electronics wants to
    explore the relationship between the number of employees
    who assemble a certain product and the number of units
    produced per hour. The complete set of paired observations
    follows:
    No. of 9 5 12 8 4 2 10 6
    Employees
    Production(units) 40 22 30 28 15 20 35 25
    Determine the coefficient of correlation using spearman rank
    correlation and test the significance at 0.05.

    Solution:

    Step 1: State the hypotheses.

    𝐻0= There is no significant correlation between the number of employees


    who assemble a certain product and the number of units produced per
    hour.
    𝜇 =0
    𝐻1= There is a significant correlation between the number of employees
    who assemble a certain product and the number of units produced per
    hour. 𝜇 ≠0

    Step 2: The level of significance and critical region.

    𝜶 =0.05, n=8, df = n-1=8-1=7 and 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = ±1.895

    Step 3: Complete the table and compute for the value of 𝝆 and

    t-test.

    Pair x y 𝑹𝒙 𝑹𝒚 D 𝑫𝟐
    1 9 40 3 1 2 4
    2 5 22 6 6 0 0
    3 12 30 1 3 2 4
    4 8 28 4 4 0 0
    5 4 15 7 8 1 1
    6 2 20 8 7 1 1
    7 10 35 2 2 0 0
    8 6 25 5 5 0 0
    Total(n=8 Σ𝑫𝟐= 10
    )
    𝝆= 1- 𝟔𝚺 𝐃𝟐
    𝐧(𝐧 𝟐−𝟏)

    𝜌 = 1- 8(86(10) 2−1)

    𝜌 = 1-

    𝜌 = 1- 60

    504
    ρ = 1- 0.1190

    𝝆 = 0.8809

    = 𝐫×√𝐧−𝟐
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 √𝟏−𝐫𝟐

    tcomputed = √1−(0.8809)0.8809×√8−22
    = √1−0.77598481
    0.809×√6
    0.8809×2.4494
    tcomputed = √0.22401519

    tcomputed tcomputed
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 = 4.5587

    Step 4: Decision rule.

    Reject the null hypothesis, if 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 is greater than the 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍

    Here, 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 > 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 4.5587 > 1.895

    So, we have enough evidence to reject 𝐻0 , so reject the null

    hypothesis.

    Step 5: Conclusion
    We can conclude that, at 0.05, significance level, there is a significant
    correlation between the number of employees who assemble a
    certain product and the number of units produced per hour. 𝜇 ≠0

    2. A used car depot wants to study the relationship between the


    age of a car and its selling price. Listed below is a random
    sample of 9 cars sold at the depot during the last 3 months.
    Car 1 2 3 4 5 6 7 8 9
    Age 2.5 4 5.5 7 6 10 5.5 4 3
    (years)
    Selling 400 300 250 190 280 120 170 320 350
    Price
    ((₱000s)
    Determine the coefficient of correlation using spearman rank
    correlation and test the significance at 0.01.

    Solution:

    Step 1: State the hypotheses.

    𝐻0= There is no significant correlation between the age of a car


    and its selling price. 𝜇 =0 𝐻1= There is a significant correlation
    between the age of a car and its selling price. 𝜇 ≠0

    Step 2: The level of significance and critical region.

    𝜶 =0.01, n=8, df = n-1=8-1=7 and 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = ±2.998

    Step 3: Complete the table and compute for the value of 𝝆 and

    t-test.
    Car x y 𝑹𝒙 𝑹𝒚 D 𝑫𝟐
    1 2.5 400 9 1 8 64
    2 4 300 6.5 4 2.5 6.25
    3 5.5 250 5 6 1 1
    4 7 190 3 7 4 16
    5 6 280 4 5 1 1
    6 10 120 1 9 8 64
    7 8 170 2 8 6 36
    8 4 320 6.5 3 3.5 12.25
    9 3 350 8 2 6 36
    Total(n=9) Σ𝑫𝟐= 236.5
    𝒎𝟏= 2 (as 4 years repeated 2 times)
    𝟏(𝒎𝟏𝟑−𝒎𝟏) 𝟏(𝒎𝟐𝟑−𝒎𝟐)

    𝝆 = 1-𝟔[ 𝚺𝐃𝟐+ 𝟏𝟐 + 𝟏𝟐 + …… ]
    𝒏(𝒏𝟐−𝟏)

    1(23−2)
    12
    𝜌
    = 6[ 236.5 + ]
    9(92−1)

    6[ 236.5 + 1(8−2) ]

    𝜌 = 1- 12
    9(81−1)

    6[ 236.5 + 1(6) ]

    𝜌 = 1- 12
    9(80)

    𝜌 = 1- 9(80) 6[ 237 ]
    𝜌 = 1-7201422

    𝜌 = 1- 1.975

    𝝆 = 0.975
    = 𝐫×√𝐧−𝟐
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 √𝟏−𝐫𝟐

    tcomputed

    = √1−0.9506250.975×√7

    tcomputed 0.975×2.6457
    =
    √0.049375

    tcomputed tcomputed
    𝐭𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐝 = 11.6089

    Step 4: Decision rule.

    Reject the null hypothesis, if 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 is greater than the 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍

    Here, 𝒕𝒄𝒐𝒎𝒑𝒖𝒕𝒆𝒅 > 𝒕𝒄𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 11.6089 > 2.998

    So, we have enough evidence to reject 𝐻0 , so reject the null

    hypothesis.

    Step 5: Conclusion
    We can conclude that, at 0.01, significance level, there is a significant
    correlation between the age of a car and its selling price. 𝜇 ≠0

    Thank you

    You might also like