(2 points) You are asked to organize a meeting on the topic of developing a marketing database to support the companys business objectives. Which of the following topics should be on the agenda for the meeting? Question 1 options: All of these items Legal considerations about what is permissible or not permissible about the data Technical considerations such as hardware and software. None of these items Competitive considerations such as comparisons vs. peer companies (2 points) Your CMO is an expert in database marketing and asks you to make sure that the company has high-quality House data to use in database marketing. House data refers to: Question 2 options: Data about the homes of your customers Economic data about trends in real-estate All of these None of these A way to differentiate customers that own vs. rent homes (2 points) Imagine that you work in the airline industry and are analyzing customer data. Which of the following would be a useful new variable for identifying customers for a Valued Frequent Flyers loyalty program? Question 3 options: A ratio variable that calculates the dollar amount per trip None of these A logic-counter variable that counts the number of different airports a customer has flown to/from A variable that measures the average number of days between trips All of these (2 points) You want to conduct an analysis to see if response rates by Age Groups (coded as <20, 20-30, 30-50, 50+) also differ by Gender (coded as M, F in the database). The appropriate analytic technique to examine this is: Question 4 options: None of these A cross tabulation of response rates by Age x Gender All of these A ratio variable created from Gender and Age A logic counter variable involving the number of Males in the database (2 points) Your database analyst asks you why there are decoy records in the database. Which of the following reasons should be part of your explanation? Question 5 options: None of these For monitoring and tracking how those records are used To always have at least one record in the database All of these To protect privacy information, so that real records need not be extracted or revealed (2 points) A match-code could be used as a value for a Primary Key field in a relational database table Question 6 options: True False (2 points) You have been asked to take over an existing relational database, so you examine one of the tables. You notice that one of the fields contains the same value for all records. Which of the following might you note in terms of the database design Question 7 options: Repetition and redundancy of information within a table is a poor design choice because it is an inefficient use of space Repetition and redundancy of information within a table is a good design choice because it prevents information loss As long as SQL code can be written against the database, the table structures don't matter. All of these None of these (2 points) You are asked to expand the design of an existing relational database system by incorporating a new table of additional information. You notice that in this new table, none of the existing fields contain unique values and so none of the fields can be used as a primary key. A good design decision would be to Question 8 options: All of these Leave the table alone and do not relate it to any other tables Break up this and all other tables so that primary keys can "emerge" and redo the relationships among tables None of these Add a new field to this table to serve as the primary key for the table (2 points) Which of the following is not a valid SQL statement? Question 9 options: SELECT * FROM jobs_table; None of these All of these PROC REG table=jobs (value='Vice President', salary=Y); SELECT title SUM(salary) FROM jobs_table GROUP BY title; (2 points) SnowSQL is the SQL language used for what database platform? Question 10 options: Microsoft database None of these ve database Snowflake database Any of these (2 points) Which of the following characteristics are true about MapReduce analytics? Question 12 options: It is a mapping technique that converts Geographic/Spatial coordinates to their geographic areas on a map It is an analytic approach where data and processing are distributed across many computer nodes running in parallel It is a way to reduce several records of multiple customers if they have the same household address All of these None of these A random sample and a representative sample are equivalent to each other Question 14 options: True False (2 points) You have built a simple regression model, and during a presentation of the model, the Chief Analytics Officer asks what is the model's "coefficient of determination"? Question 17 options: He is asking for the R-squared value None of these All of these He is trying to determine if the model is a good fit He is asking for the amount variance explained by the model (2 points) If there is a low or very weak correlation between Customer Satisfaction (Y), and Amount Spent (X), then a simple regression model using amount spent to predict customer satisfaction will not have much explanatory power. Question 18 options: True False (2 points) The difference between multiple regression and simple regression is: Question 19 options: In simple regression, there is a single predictor variable (X), but in multiple regression there can be more than one predictor (X's) In simple regression, there is a single Y variable that you are trying to predict, but in multiple regression there can be multiple measures (Y's) All of the above In simple regression, there is only one equation to summarize the relationship, but a multiple regression model is can have more than one equation. None of the above (2 points) Imagine that you are looking at a table summarizing the results of a multiple regression model. The largest coefficient value of a predictor means that variable is the most important predictor for Y Question 21 options: True, because it will have the largest effect on the predicted value, Y. False because the coefficient has no effect on the calculation of the predicted value, Y True, but only if the coefficients for the predictors are standardized coefficients All of these None of these (2 points) You run a snack shop at the beach and you find a perfect positive correlation between the weather (TEMPERATURE) and the revenue you receive from sales of drinks (REVBEV). So based on the weather forecast for the upcoming week you can predict how much revenue you'll get. Does this mean that you can also look at revenues from the past week and know (predict) what were the temperatures? Question 22 options: Yes, if you convert TEMPERATURE and REVBEV to their z-score values, then you can use the same simple regression equation both ways. None of these All of these Yes, but only if the values of TEMPERATURE and REVBEV in your regression equation use values within the range that you calculated the correlation from Yes, if you have two separate simple regression models to predict TEMPERATURE from REVBEV, and to predict REVBEV from TEMPERATURE (2 points) You can hire one database analyst, and you are interviewing three candidates for the position. Each has said they had experience handling missing data and you ask candidates what they did. Candidate Zora says she plugged missing values with 0's. Candidate Maddy says she plugged missing values with the Median. Candidate Reggi says she used multiple regression to impute values that were missing. On the basis of their answers, who is the best candidate? Question 24 options: None of these. Keep interviewing Maddy Reggi Zora Any one of these, they're all equally good. (3 points) Please rate this exam (Free points, as any answer is correct) Question 25 options: Much easier than expected Easier than I expected About what I expected More difficult that I expected Much more difficult than I expected
