Data class

Using the decision variables NO and NB, write down a linear algebraic expression for the following constraint: "The amount of fruit juice used in production during the coming week cannot exceed 15 gallons." Q4

(0.03* NO) + (0.02* NB) ≤ 15

What is the sample mean of RD's profit values corresponding to these competitor's bid values - 10.5, 12.3, 11.4, 13.8, 12.5, in millions of $? Q5

.3

Use Excel on Windows to set up and run a simulation of RD's profit using n = 100 simulation trials and the random seed of 123. Based on the results of this simulation, what is the estimate for the expected value of RD's profit, expressed in $ millions? Choose the closest value from the choices below and remember that if you are using Excel on Mac or Google sheets, your answers could be slightly different from the ones obtained using Excel on Windows as the random number generation algorithms could be different on those systems. Q5

.345

Suppose that Excel's Random Number Generation tool has generated the following sequence of five random values from the normal distribution with mean of 12 and standard deviation of 1: 10.5, 12.3, 11.4, 13.8, 12.5. These values reflect five random values of the competitor's bid, in millions of $. What is the sample mean of these random values? Q5

12.1

How many gallons of fruit juice will the production plan NO = 250 and NB = 350 require? Q4

14.5

Here is a table of the number of downloads observed in each of the past 20 periods of observation for a new pop song posted on MusicTunes in the App store. Using the Moving Averages of "5" method, i.e., MA(5), calculate the moving averages for periods 6 through 20. Now calculate the errors in your forecasts. What is the Mean Squared Error from the data? Q1 Period: (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20) Demand: (1, 39, 19, 5, 97, 44, 49, 95, 46, 56, 3, 90, 2, 19, 66, 48, 11, 92, 99, 86)

1400

Irrespective of whether or not the plan in Question 1 is feasible, what is the total revenue that FL would earn under this plan, in $? H3

177,408

Solve FL's optimization problem using Excel and Solver. What is the optimal objective function value? H3

178,652

What is the algebraic expression for the revenue, in $, as a function of decision variables A1, A2, A3, M1, M2, M3? H3

19*A1 + 19.2*A2 + 19.4*A3 + 28.8*M1 +28.16*M2 + 28.48*M3

Consider the demand distribution for the store per the data below. Find the standard deviation of this demand distribution. Scenario 1: D1 = 10 with probability p1 = 0.2 Scenario 2: D2 = 20 with probability p2 = 0.2 Scenario 3: D3 = 40 with probability p3 = 0.2 Scenario 4: D4 = 60 with probability p4 = 0.2. Scenario 5: D5 = 80 with probability p5 = 0.2 H1

25.61

The Surf-n-Turf store in Sheboygan sells surfboards. The seasonality factors for the demand for one particular model of surfboard over 12 months of the year are as follows in the chart below. The de-seasonalized forecast for demand in the month of July is 1433 surf boards. What is the correct seasonalized forecast for the actual demand in July? H2 Month Seasonal Factor January. 0.6 February. 0.7 March. 0.9 April 1.1 May 1.5 June. 1.6 July 1.8 August 0.9 September 0.7 October 0.8 November 0.7 December 0.7

2579.4

In general, how good is data analysis in predicting variability in job performance? How much variation in performance does it generally predict in candidates? Select the most appropriate option. Q6

30% to 40%

Below is a picture of an excerpt from an Excel file set up to model ZI's profit maximization problem, with the number in cell D3 hidden. The formula in cell E7 is =SUMPRODUCT(C$3:D$3,C7:D7). What is the value in cell D3? Q4

300

Examine the data from the above question. Using the Moving Averages of "3" method, i.e., MA(3), calculate the moving averages for periods 4 through 20. Now calculate the errors in your forecasts. What is the Mean Absolute Percentage Error calculated from the data? Choose the closest numerical answer. Q1 Period: (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20) Demand(1, 39, 19, 5, 97, 44, 49, 95, 46, 56, 3, 90, 2, 19, 66, 48, 11, 92, 99, 86)

300%

Suppose that ZI decides to make 250 boxes of Ortaköy and 350 boxes of Beyoglu. What is the weekly profit that it will earn? Q4

3105

Solve the optimization problem using Excel and Solver. What is the optimal objective function value? Remember that given the relatively small size of the market for Turkish Delight, ZI cannot sell more than 250 boxes of Ortaköy(O) and 400 boxes of Beyoglu (B) in the coming week. Q4

3285

Using the same Excel sheet, what value will appear in cell E8 if the formula in cell E7 is copied and pasted into the cell E8? Q4

35

Company W owns pharmacies/drug stores in several US cities. Demand for Flu shots at a store in center-city Philadelphia, can be modeled by a random variable that takes the following values in five different scenarios that occur with the following probabilities: Scenario 1: D1 = 10 with probability p1 = 0.2 Scenario 2: D2 = 20 with probability p2 = 0.2 Scenario 3: D3 = 40 with probability p3 = 0.2 Scenario 4: D4 = 60 with probability p4 = 0.2 Scenario 5: D5 = 80 with probability p5 = 0.2 What is the mean of this demand distribution? H1

42

Write down an algebraic expression for ZI's weekly profit, in $, as a function of the decision variables NO and NB. Q4

5.0* NO + 5.3* NB

Assume you are the retailer who orders goods at 3 talers an item from the manufacturer and sells it in stores at 12 talers an item to customers. You decide your order quantity Q and place the ordered units on shelves before the uncertain demand is realized. Consider the following scenario: You decide to buy Q=100 items and the customer demand turns out to be 87 items. Calculate the profits accrued in this scenario. H1

744

What is an action? Select all options that apply. Q7

A choice that impacts the goal

Which of the following are threats to Amazons's use of advanced predictive shipping? Q9

A lack of local distribution centers

Using these decision variables, what is the algebraic expression for the constraint of "the total number of fruit cartons shipped in Truck 1 cannot exceed this truck's capacity"? H3

A1 + M1 ≤ 2700

When would descriptive and predictive results need additional analysis? Select the most appropriate option. Q7

All are correct (When the firm can make a choice of different actions to take, When there are strategic consumers involved, When there are multiple explanations to the same data we observe, When there is competition involved)

Why is it important to consider strategic interaction? Select the most appropriate option. Q7

All of the above (Strategic interaction can affect the validity of your model, You need to ensure you are asking the right question, It can affect the recommendations you make)

Why is turnover costly K1

All of the above (Training costs, impacts on colleagues, hiring costs)

Which of the following is GPS-based tracking likely to enable firms to do? K1

All of these are correct (Lower churn rates, Raise CLV (Customer life value), provide targeted advertisements at optimal times)

Which of the following is GPS-based tracking likely to enable firms to do? Q9

All of these are correct (Lower churn rates, Raise CLV, Provide targeted advertisements at optimal times)

What is NOT a weakness of conducting simple comparisons of attrition rates across groups? Select the most appropriate option. H5

Attrition rates (% of workers who leave) is a poor measure of turnover

Maxi-min strategy: Q8

Chooses actions that maximize the minimum outcome

The 3 building-block of decision tree do NOT include: Q8

Circular decisions

What is one way to collect data that will help predict internal mobility success? Select the most appropriate option. Q6

Collect data on characteristics of employees across the company and match performance predictors to specific positions

What is the main drawback of using the Moving Averages forecasting method for forecasting data that shows an increasing trend? H2

Moving average forecasts lag behind the trend in the data.

Why did Bidwell's 2011 study of an investment firm indicate that internal hiring might be a better investment than outside hiring? Select all options that apply. Q6

Outside hires take about 3 years to perform at the level of internal hires, Internal hires tend to be paid less than outside hires

What information would you want to compare different methods for filling jobs internally (e.g. posting versus sponsorship)? Select the most appropriate option. H5

Performance before and after moves, as well as how the job was filled

"people continue to get promoted until they hit a job they cannot do" K1

Peter Principle

Which of the following is not a method used to track customers across webpages? Q9

Pop-up Advertising

which of the following is not a method used to track customers across webpages? K1

Pop-up advertising

In which of these situations would it be more appropriate to use a probability model rather than a regression/data-mining approach? Q3

Predicting when the customer will make her next purchase

Sales of a product is a dependent variable for an organization. They have conducted two different regression analyses for two different independent variables. Using price as the independent variable, the regression model has an R2=0.91. If, instead, the independent variable is the number of times an advertisement displayed on television, R2 = 0.43. Which of the two independent variables leads to a better prediction of future sales? Q3

Price of the product

What feature should you use if you want to generate 10 numbers with normal distribution with mean of 20 and standard deviation of 5? H4

Random number generation

Why does the "RFM" rubric present the three key measures (recency, frequency, and monetary value) in that order? Q3

Recency trumps frequency, and monetary value is far less important than both of them

Among the explanations below, which one is a reason to favor a probability model over a regression-like (e.g., data-mining) model for long-run projections of customer behavior? Q3

Regression-like models are fine for a one-period-ahead prediction, but not beyond that horizon

When we refer to a "cohort," we are talking about a group of customers who: Q3

Share similar acquisition characteristics (e.g., time of acquisition)

What does online retargeting do? Select the most appropriate option. Q7

Shows ads to people after they visited a specific website

What technology is Comcast using to improve customer satisfaction? Q9

Software that reads customer intonation

What technology is Comcast/Xfinity using to improve customer satisfaction? K1

Software that reads customer intonation

Which of the following is NOT a reason why it is useful to understand the main drivers of turnover within a company: Select the most appropriate option. H5

To identify those jobs where the company should focus its retention efforts

Which software is used in class to solve simulation problems? H4

TookPak / Data Analysis (Excel Add-in)

Which of the following is the biggest challenge to solving the "advertising attribution problem"? Q9

Tracking customers across digital properties is difficult.

What is the difference between structured and unstructured interviews? Q6

Unlike unstructured interviews, structured interviews are designed to assess specific attributes

Referring to the dataset (and model) we covered extensively, how would these two customers (both acquired in 1995) compare to each other, in terms of their expected future purchasing? Q3 1995 1996 1997 1998 1999 2000 2001 John. 1. 0. 0. 0. 1. 1. 1 Venus. 1. 1. 1. 1. 0. 0. 1

Venus would likely be more valuable

Why does it matter to know how a demand curve was generated? Select the most appropriate option. Q7

We may give a different recommendation for different models

Prescriptive analytics answers the following question H6

What can we do to change an outcome in our favor?

When could you be confident that the effect of a given intervention (e.g. new incentive system) on performance was causal (i.e., the intervention caused a change in performance)? Select the most appropriate option. H5

When assignment to the intervention was unrelated to any factors that might influence performance

When does including controls in multivariate regressions not solve problems of omitted variable bias? Select the most appropriate option. H5

When important potential omitted variables cannot be measured with precision

When does maximizing revenue also maximize profit? Select all options that apply. Q7

When there is no cost to the product, When the marginal cost is zero

WTP H6

Willingness to pay

reverse causality K1

Y causes X instead of X causing Y

Suppose that FL plans to put the same number of cartons for each fruit on each truck. In other words, suppose that FL sets A1 = A2 = A3 = 1300 and M1 = M2 = M3 = 1200. Is this a feasible plan? H3

Yes

How is multivariate regression helpful for predicting performance? Select all options that apply. Q6

You can analyze multiple characteristics, or predictors, at the same time, You can determine how strongly a characteristic is predictive of performance, it gives you a way to quantify a variable

Which of the following is a weakness of using multivariate regression for predicting which hires will perform best? Select the most appropriate option. H5

You only have data on the people that you actually hired

in people analytics, the 3 components of the staffing cycle are 'hiring', 'internal mobility and career development', and __________. K1

attrition

the three building blocks of a decision tree K1

decision nodes, event nodes, outcomes

collect data on consumer behavior and make predictions/forecasts using trends and seasonality K2

descriptive analytics

the three pillars of data analytics K2

descriptive, predictive, perscriptive

Why is there only a moderate correlation between job satisfaction and turnover? Select the most appropriate option. H5

Because levels of job satisfaction only appear in some of the paths that lead to turnover

Why might we see performance improve after training, even if the training is completely ineffective? Select the most appropriate option. H5

Because people are more likely to go to training when their performance dips

What does a "BTYD" model refer to? Q3

Buy 'Til You Die

When comparing new hire predictions, what are the issues Bidwell (2011) highlights that should be acknowledged or addressed? Select all the options that apply. Q6

Ensuring you are comparing similar candidates, or comparing apples to apples, Disentangling influences, Accounting for selection, or who was actually hired

What does a model do? Select the most appropriate option. Q7

Explains the relationship between actions and parameters to the goal

Which of these statements is most aligned with our assumption(s) about randomness when it comes to modeling/explaining customer behavior? Q3

Customers are not truly random but appear to be "as if" random from an outsider observer's perspective

Moving averages (MA) method is used for K1

Forecasting

Three parts to solve a problem K1

Goal, actions, and model

Which of the following is a critical requirement for predicting who will perform well following promotion? Select the most appropriate option. H5

Data on the workers who enter their jobs directly by hiring versus being promoted

Which one of these is an example of a tradeoff? Select the most appropriate option. Q7

Data on the workers who enter their jobs directly by hiring versus being promoted

The first step in Making Decisions in High Certainty Environments: H4

Decide on reward and risk measures

Points at which decision-maker must decide on an action K1

Decision node

When targeting customers for optimal marketing, which of the following rank ordering of customers from highest to lowest is most appropriate to determine which customers to target? Q9

Highest to lowest marketing effectiveness

when targeting customers for optimal marketing, which rank ordering of customers from highest to lowest is appropriate? K1

Highest to lowest marketing effectiveness

Which type of data provides the most granular level of information about a given individual's customer behavior? Q9

Household-level scanner data from the store

Why we say that hiring may be the most important part of people analytics? Q6

If you hire people who are a good fit for the company, then you have to worry less about performance management

Which of the following statements is a true statement about the Newsvendor model? H1

In the Newsvendor model, the ordering decision is made before seeing the customer demand.

Why is it important to account for selection when seeking to find performance indicators by analyzing your current employees? Select the most appropriate option. Q6

It is not the same pool as general applicants.

What is the goal of prescriptive analysis? Select the most appropriate option. Q7

Make a recommendation on an action that will optimize a goal

MR = MC K1

Marginal revenue = Marginal cost

Which one of the following is an example of a goal/objective? Select the most appropriate option. Q7

Maximize the quantity of a product sold

A data set has 20 points drawn from a particular demand distribution. Find the descriptive sample statistics: namely, the sample mean and standard deviation. Q1 (78, 66, 95, 87, 57, 97, 114, 44, 74, 85, 95, 88, 69, 86, 50, 70, 90, 43, 61, 80)

Mean = 76.45 and standard deviation = 18.96

Which of the following is typically marked with a circular icon? Q8

event node

Which type of data provides the most granular level of information about a given individuals customer behavior? K1

household-level scanner data from the store

Additional revenue by selling one more item H6

marginal revenue

Risk seeking strategy when analyzing decision tree K1

maxi-max

nps refers to K2

net promoter scores

ZI is considering setting its production for the next week at the maximum levels that the market will bear - in other words, it is considering a production plan NO = 250 and NB = 400. Is this plan feasible? Q4

no

In general, which hiring activity has the strongest relationship with performance? Select the most appropriate option. Q6

prior work samples

Which tool can we use to run simulations with decision tree on Excel? Q8

solver add-in

Is G.P.A. a valid predictor of performance? Q6

sometimes

Which of the following is the biggest challenge to solving the "advertising attribution problem" K1

tracking customers across digital properties is difficult

Which of these means of assessing candidates generally has the lowest correlation with subsequent performance? Select the most appropriate option. H5

unstructured interviews

In a data set with 2500 points, the sample mean was calculated to be 3980, and the sample standard deviation was calculated to be 1000. Assuming that the data were generated from a random output that was distributed according to a Normal distribution, what is the mean and standard deviation that you would use for predicting the next sample. H2

µ = 3980 σ=1020