Homework answers / question archive / After you run a campaign and get back purchase decisions, you'd like to figure the probability of a customer from your database ordering based upon four continuous variables

After you run a campaign and get back purchase decisions, you'd like to figure the probability of a customer from your database ordering based upon four continuous variables

Business

Share With

After you run a campaign and get back purchase decisions, you'd like to figure the probability of a customer from your database ordering based upon four continuous variables. Which of the following techniques will you use? A. Correlation B. Chi-Squared Analysis C. Logistic Regression D. Multiple Linear Regression You might have been asked to choose the closest to correct model for the response variable: (Hint: you'll need to recode Response from Y/N to something else) Using the p-values of the coefficients in the result we can reduce the linear equation to a simpler form: Data for Linear Regression final exam.docx • A. Response = .818 - .00785 * Age • B. Response = 3.98-.05 * Height • C. Response = 4.28-.051 * Height • D. No change, all coefficients are significant • E. Response = 5 -.0644 * Age 54 KB In BigClass, The correlation between AGE and WEIGHT is? BigClass.xlsx 37 KB • A. Cannot be determined • B. No correlation • C. Negative • D. Positive Propose a multiple-choice question covering the materials from the course and provide at least four answers to it. Very good questions will earn 2 points bonus. Why is NYU SPS not a good representative sample of New York City? • A. There are 40 other colleges within an hour drive of East 42nd street • B. New York is diverse, made up of many types of people of all ages and college graduation status • C. There are many other ethnic groups in New York • D. NYU and New York City are made up of many types of students We need to reach people living in a specific zip code in Brooklyn, primarily male, Hispanic, 35 to 54. What would a representative sample of this population look like? • A. Older than our target audience • B. Younger than our target audience • C. The same as our target audience • D. A mixture of male and female Analyzing the attached chart what kind of correlation do you see? Input for Q24.docx 444 KB • A. Positive correlation • B. Negative Correlation • C. Negative to positive correlation • D. No correlation • E. Positive to negative correlation You have run a marketing promotion that, in the past, has returned 14% response rate from the targeted group. You've recently decided to tweak the campaign to see if you get a higher response rate, so you run a campaign using both versions of the promotion, and evaluate it using chi-squared analysis. Of 4790 people promoted under the new model, 730 responded, a Blank 1. Fill in the blank, read surrounding text. % response rate when rounded to the nearest whole percentage; meanwhile 750 of 5210 people promoted with the old promotion responded, giving you a Blank 2. Fill in the blank, read surrounding text. % response rate, when rounded to a whole percentage. You decide that this Blank 3. Fill in the blank, read surrounding (write is or is not) a statistically valid result. The primary goal in establishing a marketing database is: • A. Using technology to organize customer accounts • B. Segmenting customers according to geographic area • C. Helping the organization achieve an objective such as increasing profits by tracking campaigns • D. Developing fast technology • E. Segmenting customers according to geographic codes CMO Grace Liu wants to aggregate data together because individual data points are too few by themselves to generate meaningful information. What is the CMO creating? • A. Logic counter variables text. • B. Univariate variables • C. Cross tab variables • D. Longitudinal variables Using the attached chart, the tab for question 23, if Chou En Lai figured out that the goal of the promotion is to generate a 6% profit after overhead and the resulting response rate had to be 3.7%, which names can Chou promote profitably? Input for Q22,23,25,26,28,29,30.xlsx • A. None • B. 51 to 60 • C. 41 to 50 • D. Over 61 85 KB Choose the closest to correct model for the response variable: (Hint: you'll need to recode Response from Y/N to something else) Data for Linear Regression final exam.docx • A. Response = 4.28 -.051 *Height - .008 * Age • B. Response = b+ m1 * Age +m2 * Height • C. Response = 5 +.064 *Age - .00007 * Height • D. Response = 5 -.0644 * Height - .00007 * Age 54 KB Data for Part B Consider the following customer database. They were recently promoted and the company accumulated their responses to the campaign as shown. They want to use this data to promote again but only to the most responsive customers. Use Excel and create a linear response model to answer the following questions. Customer Name Alan Bob Jessica Elizabeth Hilary Fred Alex Margot Sean Chris Philip Catherine Amy Erin Trent Preston John Nancy Kim Laura Height Age 70 72 65 62 67 69 65 63 71 73 75 70 69 68 72 68 64 64 72 62 39 21 25 30 19 48 12 51 65 42 20 23 13 35 55 25 76 24 31 29 Actual Response N Y Y Y Y N Y Y Y N Y N N Y N N N Y N Y name age KATIE LOUISE JANE JACLYN LILLIE TIM JAMES ROBERT BARBARA ALICE SUSAN JOHN JOE MICHAEL DAVID JUDY ELIZABETH LESLIE CAROL PATTY FREDERICK ALFRED HENRY LEWIS EDWARD CHRIS JEFFREY MARY AMY ROBERT WILLIAM CLAY MARK DANNY MARTHA MARION PHILLIP LINDA KIRK LAWRENCE 12 12 12 12 12 12 12 12 13 13 13 13 13 13 13 14 14 14 14 14 14 14 14 14 14 14 14 15 15 15 15 15 15 15 16 16 16 17 17 17 gender F F F F F M M M F F F M M M M F F F F F M M M M M M M F F M M M M M F F M F M M height weight 59 61 55 66 52 60 61 51 60 61 56 65 63 58 59 61 62 65 63 62 63 64 65 64 68 64 69 62 64 67 65 66 62 66 65 60 68 62 68 70 95 123 74 145 64 84 128 79 112 107 67 98 105 95 79 81 91 142 84 85 93 99 119 92 112 99 113 92 112 128 111 105 104 106 112 115 128 116 134 172 Questions 22 and 23 % of Sample Age Number 30 and under 1.600,00 0,1038961 31 to 40 4.000,00 0,25974026 41-50 2.500,00 0,16233766 51-60 3.600,00 0,23376623 61 and over 1.700,00 0,11038961 No age avail 2.000,00 0,12987013 total 15.400,00 1 Number of Orders 36 99 91 29 46 51 352 Questions 22 and 23 Response Rate Index to Total 2,25% 2,48% 3,64% 0,81% 2,71% 2,55% 2,29% 0,98 1,08 1,59 0,35 1,18 1,12 1,00 Q25, 26 Customer Address Chu Aggata Andres Beverly Beijing Warsaw Rome Marblehead Total Promotions 15 25 17 29 Q25, 26 Ratio Orderts to Total Orders Promotions 5 12 5 4 33% 48% 29% 14% Number of Past Purchases 0-1 Purchases 2-4 Purchases 0-3 Months Ago RR = 5.34% (106) R Ord = 285 O C1 T RR = 7.54% (150) R Ord = 361 O Tot = 5,337 Tot = 4,789 C6 T RR = 11.23% (224) R Ord = 76 5-10 Purchases Tot = 677 O C11 T RR = 14.71% (293) R Ord = 20 11+ Purchases TOTAL O C16 T RR = 6.78% (135) R Ord = 742 O Tot = 10,939 T Tot = 136 s s Q28,29,30 Last Purchase Date 0-3 Months Ago 3-6 Months Ago 6-9 Months Ago 9-12 Months Ago 12+ Months Ago TOTA RR = 5.34% (106) RR = 4.58% (91) RR = 3.75% (75) RR = 2.98% (59) RR = 1.45% (29) RR = 3.37% Ord = 285 Ord = 383 Ord = 428 Ord = 488 Ord = 139 Ord = 1,723 Tot = 11,420 C3 Tot = 16,391 C4 Tot = 9,568 Tot = 5,337 C1 Tot = 8,354 C2 C5 Tot = 51,07 RR = 7.54% (150) RR = 6.57% (131) RR = 4.98% (99) RR = 4.35% (87) RR = 2.79% (56) RR = 4.56% Ord = 361 Ord = 945 Ord = 1,098 Ord = 1,314 Ord = 721 Ord = 4,439 Tot = 14,376 C7 Tot = 22,039 C8 Tot = 30,203 C9 Tot = 25,838 C10 Tot = 97,24 RR = 11.23% (224) RR = 9.44% (188) RR = 6.45% (128) RR = 5.45% (109) RR = 4.48% (89) RR = 5.57% Ord = 76 Ord = 801 Ord = 1,418 Ord = 809 Ord = 3,296 Tot = 2,033 C12 Tot = 12,426 C13 Tot = 26,018 C14 Tot = 18,051 C15 Tot = 59,20 RR = 14.71% (293) RR = 11.46% (228) RR = 8.82% (176) RR = 7.01% (140) RR = 6.34% (126) RR = 7.30% Ord = 20 Ord = 792 Ord = 1,448 Ord = 763 Ord = 3,100 C17 Tot = 8,981 C18 Tot = 20,654 C19 Tot = 12,036 C20 Tot = 42,47 RR = 6.78% (135) RR = 6.28% (125) RR = 5.68% (113) RR = 5.01% (100) RR = 3.71% (74) RR = 5.02% Ord = 742 Ord = 1,597 Ord = 3,119 Ord = 4,668 Ord = 2,432 Ord = 12,55 Tot = 10,939 Tot = 25,435 Tot = 54,867 Tot = 93,266 Tot = 65,493 Tot = 250,0 Tot = 4,789 Tot = 677 Tot = 136 C6 Ord = 192 C11 Ord = 77 C16 Tot = 672 12+ Months Ago TOTAL = 1.45% (29) RR = 3.37% (67) = 139 Ord = 1,723 = 9,568 C5 Tot = 51,070 = 2.79% (56) RR = 4.56% (91) = 721 Ord = 4,439 = 25,838 C10 Tot = 97,245 = 4.48% (89) RR = 5.57% (111) = 809 Ord = 3,296 = 18,051 C15 Tot = 59,206 = 6.34% (126) RR = 7.30% (145) = 763 Ord = 3,100 = 12,036 C20 Tot = 42,479 = 3.71% (74) RR = 5.02% (100) = 2,432 Ord = 12,558 = 65,493 Tot = 250,000 Q13-34 From the attached Excel spreadsheet go to the Excel tab for Q25 and Q26. Determine which person had the highest order to promotion ratio: Input for Q22,23,25,26,28,29,30.xlsx • A. Beverly • B. Chu • C. Andres • D. Aggata 85 KB Use the BigClass.xls file associated with this exam. Open it in Excel and compute the correlations asked for. What is the correlation between the AGE and HEIGHT variables? BigClass.xlsx • A. .60826 • B. .46318 • C. .70917 • D. none of these 37 KB Which of the following is not a valid reason for purging or removing customers from your marketing database? • A. Unable to confirm address for past 2 to 3 years • B. Inactive for the past three years • C. Customer only ordered during the past 6 month promotional period • D. Non-deliverable or invalid mailing address From the Excel spreadsheet. tab for question 28, 29, 30: what is the cumulative response rate for those people with a response rate of 7.5% or greater? Input for Q22,23,25,26,28,29,30.xlsx • A. 29% • B. 63.2% • C. 5% • D. 9% 85 KB In BigClass, Why is it not possible to compute the correlation between AGE and GENDER? BigClass.xlsx 37 KB • A. AGE is an integer variable • B. GENDER is a categorical variable • C. You need to use Chi Squared • D. It IS possible to compute the correlation between these two variables in their current format Using the attached chart, go to the tab for question 22: assuming the breakeven for hot fire chilies in a jar is a response rate of 2.50%, which names could a product manager profitably promote? Input for Q22,23,25,26,28,29,30.xlsx • A. 41-50, 61 and over, no age available 85 KB • B. No age available and 31 to 40 • C. 30 and under and 31 to 40 • D. 51 to 60 and 41 to 50 • Working with categorical data requires some special handling techniques, of which we've used a few. Which of the following techniques are valid when used correctly with categorical data? A. Correlation B. Linear Regression C. Chi-Squared D. Logistic Regression A key assumption of CRM is: • A. Retaining existing customers is usually more profitable than acquiring new ones • B. A focus on immediate sales volume leads to better CRM • C. New customers have lower overall costs • D. Constantly contacting good customers always yields lower costs What is the best way to do an Nth select on Excel so you get about N rows but every row has an equal chance of being selected? • A. Assign a random number between 1 and N inclusive to each row, and choose every row that gets a 1 • B. Choose a random row and then every Nth row after that • C. Choose Every Nth row from a table • D. Get drunk and delete rows until you have N rows left In BigClass, the correlation between AGE and WEIGHT makes sense since BigClass.xlsx 37 KB • A. There is no discernible correlation between these two variables • B. The older the teenager is, the more likely his/her parents have run out of money to feed him/her • C. The older the teenager is the heavier his/her body probably is • D. The older the teenager is the more attention s/he pays to diet Suppose, instead of 95% confident as in Question (Go to the sample size calculator: https://www.surveysystem.com/sscalc.htm You want to do a representative sample for a political poll on the population of China, given as 1400000000, with 95% certainty that your poll result will be within 2% correct [your confidence interval]. How large will your sample size be?) you want to be 99% confident. How large does your sample size have to be? • A. 1368 • B. 6321 • C. 2401 • D. 4160 Which of the following is a standard database maintenance routine? • A. House-holding the customer file • B. Deduping the customer file • C. All of the selections • D. Purging old customer records CEO Launa Lee is concerned about a biased result in her analysis, thus she needs to do what? • A. Pull names randomly • B. Find people who want to be taken off her list • C. Eliminate anyone who ever responded via social media • D. Pull names that are independent of the final result Based on the Ratio Variable in the attached Excel spreadsheet under tab Q25, 26, who would you be least likely to promote in the future based on just orders? Input for Q22,23,25,26,28,29,30.xlsx • A. Chu • B. Andres • C. Aggata • D. Beverly 85 KB In BigClass, The correlation between the AGE and WEIGHT variables can be considered a strong correlation. BigClass.xlsx 37 KB True False Using the attached spreadsheet, go to tab for question 26 for this question. For reasons you can’t believe, CMO Martin wants you to promote the two people who have the highest combined orders. Which two people would that be? Input for Q22,23,25,26,28,29,30.xlsx • A. Beverly and Andres • B. Aggata and Andres • C. Aggata and Chu • D. Aggata and/or Andres or Chu 85 KB Go to the sample size calculator: https://www.surveysystem.com/sscalc.htm You want to do a representative sample for a political poll on the population of China, given as 1400000000, with 95% certainty that your poll result will be within 2% correct (your confidence interval). How large will your sample size be? • A. 1644 • B. 1477 • C. 1368 • D. 2401 From the attached chart in the tab entitled Q28, 29, 30, compare cells C1, C6, C7, C13, C14, C19, C20. Determine which cell has the highest index number. What does it mean? Select the best answer below Input for Q22,23,25,26,28,29,30.xlsx • 85 KB A. Cell C7 has a 31% more likely chance of buyers buying than any other segment in the 3 to 6th month ranges • B. Cell C6 has a 50% greater chance of responding than the overall population • C. Cell C6 has a 50% greater chance of responding vs the C11 segment • D. Cell C18 has a 128% greater chance of responding than all 12+ months buyers Which of the following is most analagous to cloning your best customers from your marketing database? • A. Facebook Analytics • B. Google Analytics • C. Spokeo Search • D. Google Mail Marketing • E. Facebook Lookalike Modeling Consider the attached customer database. They were recently promoted and the company accumulated their responses to the campaign as shown. They want to use this data to promote again but only to the most responsive customers. Use Excel and create a linear response model to answer the following questions. Are Age and Height meaningfully correlated for these customers? Data for Linear Regression final exam.docx 54 KB • A. Yes, but slightly • B. Yes, negatively correlated • C. There is not enough information to answer the question • D. No Bootstrapping is A. A method of analyzing boot sales for Zappos B. A method of repeatedly selecting random samples from too small a sample set for verification to find significant variables to build a model C. Like a historical gains fall-off chart, an acceptable method to estimate the reliability of an analysis that cannot otherwise be verified D. A method of estimating sample size when sample size is too small for analysis verification Recency, frequency and monetary value data are classified as: • A. Survey data • B. Customer contact data • C. Marketing data • D. Fulfillment data After you run a campaign and get back purchase decisions, you'd like to figure the probability of a customer from your database ordering based upon four continuous variables. Which of the following techniques will you use? A. Correlation B. Chi-Squared Analysis C. Logistic Regression D. Multiple Linear Regression • Working with categorical data requires some special handling techniques, of which we've used a few. Which of the following techniques are valid when used correctly with categorical data? A. Correlation B. Linear Regression C. Chi-Squared D. Logistic Regression Bootstrapping is A. A method of analyzing boot sales for Zappos B. A method of repeatedly selecting random samples from too small a sample set for verification to find significant variables to build a model C. Like a historical gains fall-off chart, an acceptable method to estimate the reliability of an analysis that cannot otherwise be verified D. A method of estimating sample size when sample size is too small for analysis verification