Introduction

In this Analytical Techniques for Supply Chain Management Assignment, you will be presented with two questions, each weighted at 45 marks. Questions are designed in a way that you get a chance to revisit the models and tools we have learned in class. While the supply chain context of the questions remains the same, each student has a dataset prepared for themselves to avoid high similarity rates on Turnitin. Please limit your submission to 1,000 – 3,000 words.

Please confirm that you downloaded the dataset titled your student number

(“S123456.xlsx”) before answering the questions.

Please use the discussion board for any questions you might have about this assignment.

We will answer all questions that are asked two days before the submission deadline to minimise last-minute stress and encourage timely attention to the assignment.

It has been a great pleasure to teach you these subjects. We hope you enjoy this assignment and use the techniques in the future.

Prof Dr Emel Aktas

November 2023

Disclaimer: Although the problems presented in this assignment are informed by real events, the names, characters, businesses, places, events, locales, and incidents are the products of the author’s imagination. Any resemblance to actual persons, actual businesses, or actual events is coincidental.

1     Inferential Statistics [45 marks]

The sheet titled “ Inference.xlsx” includes the delivery data captured by a major online grocery retailer through their loyalty programme as well as the results of a home delivery satisfaction survey using a sample of online grocery orders in the first five days of January 2019. The dataset contains information about the following variables:

• Customer satisfaction score (A value between 1 and 6 with 1 being not at all satisfied and 6 being fully satisfied. The score is calculated based the average rating of multiple   satisfaction questions.)

• Delivery tracking (A variable consisting of “available” where the delivery can be tracked and “Not available” where the delivery cannot be tracked)

• Number of Items in the basket (Number of items that were included in the order)

• % of perishable items in the basket ((Total value of perishable items that were included in the order/Total value of the order) x 100)

• Customer’s sensitivity to delivery cost (A value between 0 and 5 with 0 being not at all sensitive and 5 being highly sensitive)

• Customer engagement score (The sum of the number of clicks on different delivery service functions between the order and the delivery time)

• Number of previous online orders (Over the last year)

• The time between order and delivery of the last order (in hours)

• The delivery delay of more than 24 hours (A variable consisting of “no” and “yes”)

The director of supply chain is preparing a report for the board and has asked you to analyse the data and derive insights and conclusions.

a.   Build an appropriate inferential statistics model to explain the variability in customer satisfaction scores.

b.   Interpret the results you obtain from your inferential statistics model.

c.   Check that the assumptions of your model hold as applicable.

d.   How do you suggest improving your model?

2     Mathematical Modelling [45 marks]

The workload in many areas of publishing operations follows a non-uniform distribution with respect to time of day. For example, at Cranfield Design and Print Studio in Cranfield University, the number of print requests put in by customers, if plotted against the time of day, would appear to have the shape of an inverted and skewed U curve, with the peak around 3 pm For efficient use of resources, the personnel available should vary over the day. Figure 1shows an atypical workload curve and corresponding personnel requirements at different times of the day.

Figure 1 Print room workload requirement overtime

The variable capacity of an operation can be achieved effectively by employing part-time personnel. Part-timers are paid less per hour than full-time employees. Other considerations, however, may limit the extent to which part-time workers can be hired in a given part of the Design and Print Studio. The problem is to find an optimum workforce schedule that would meet personnel requirements at any given time whilst being economical. Some of the factors affecting personnel assignment are listed here:

•    Full-time employees work for 8 hours per day plus a 1-hour break for lunch.

•    Approximately half of the full-timers go to lunch between 12 pm and 13 pm, and the remaining go between 13 pm to 14 pm.

•    Part-timers work for at least 3 continuous hours but no more than 5 continuous hours per day and are not allowed a lunch break.

•    By corporate policy, part-time personnel hours are limited to a maximum of 40% of the  day’s total requirement. For example, if the total personnel hours required over a day is 100 hours, then at most 40 hours of this could be met by part-time personnel. This is a  policy generally applied to ensure high ownership of the service business.

•    The shift starts at 9 am and ends at 7 pm (i.e., overtime is limited to 1 hour). Any work left over at 7 pm is considered a backorder for the next day.