You are on page 1of 9

Assessment 1:

Individual

Course code: OMGT 1039


Student name: Nguyen Thanh Van
Student ID: s3742938
Lecturer’s name:
Dr. Zaheed Halim and
Dr. Hassan Mirzahosseinian
Table of Contents:
I. Case study 1:.............................................................................................................................2
1. Pareto analysis and Cause-and-Effect Diagram:...........................................................................2
2. The different of survey results depending on the two separate type of data: likert scale and
open-ended questions:........................................................................................................................... 4
3. Could the survey have been more useful? The further analyses can be done using collected
data and the next step is?....................................................................................................................... 4
II. Case study 2:.............................................................................................................................4
III. Reference:..............................................................................................................................9

I. Case study 1:
1. Pareto analysis and Cause-and-Effect Diagram:
 Pareto Analysis:
In accordance with Vilfredo Pareto (1895), Pareto Analysis is in any series of controlled elements, and
elected small fraction, regarding numbers of elements, always accounts for a large fraction of effect
(Goodman, 2007).
Due to the two different types of data, there are two charts of Pareto Analysis. In terms of Likert scale data,
so as to find out the root for customers' dissatisfaction, the rate for 0, 1, and 2 is chosen to appear in Pareto
analysis; thus, since the third and fourth ones are good ranking, they are not analyzed.

Pareto Analysis of Likert Scale


200 120
180

Culmulation Percentage
100
160
140
80
120
Count

100 60
80
40
60
40
20
20
0 0
1 2 3 4 5 6 7 8

Complaints

Pareto Analysis extracted from the Likert Scale

As can be seen in the chart above, the three primary causes of customer dissatisfaction are seating,
service efficiency, and catalog. To be more specific, seating occupied the highest number of defects,
accounted for 182 answers, and followed by catalog and service efficiency, amounted to 177 and 179
answers, respectively. As a result, 80% of the problems caused by 50% of the factors, according to the
Pareto principle of unequal distributions of causes and effects.

Moving to the Pareto chart of open-ended questions:


Pareto Analysis of Open-ended Questions
14 0.9

Culmulative Percentage (%)


12 0.8
0.7
10
0.6
8 0.5
Count

6 0.4
0.3
4
0.2
2 0.1
0 0
1 2 3 4 5 6 7 8

Complaints

Pareto Analysis extracted from the Open-ended questions

Regarding this Pareto chart, although the total complaints go up to the number of 49, the actual opinions
are 47 which include both positive and negative perspectives due to the two unclear complaints which
cannot be arranged in any category.

As depicted in the chart, seating and service efficiency are the biggest weight which took up to 85% and
65%, corresponding to 13 and 9 complaints in the open-ended questions. Therefore, 80% of the problems
caused by 50% of root causes which guides the efforts to enhance the process (Sarkar, Mukhopadhyay, &
Ghosh, 2013), according to the Pareto principle of unequal distributions of causes and effects.
Nevertheless, there are differences in the number of defects and the categories. In simple terms, regarding
catalog, based on 200 customer opinions, it is ranked at the third biggest cause; however, moving to the
open-ended part, relied on 47 complaints, no one complained about this category.

 Cause and Effect diagram

Cause and Effect diagram or Fishbone diagram is a graphic technique and an efficient tool to find and
analyze affecting factors on the features of work output quality (Slameto 2016). Its function is listing the
distinct root causes that can be attributed to a problem and determining the primary one.

Cause-and-Effect Diagram

The fish-bone diagram above demonstrates the five main causes that influence customer dissatisfaction
which summarized from both Likert-scale and open-ended questions. In conclusion, seating and service
efficiency are 20% of the causes generated by 80% of problems. Since Luke Forley, as the NSW
opposition leader, stated that schools and hospitals before stadiums after receiving nearly 60% of NSW
voters on rebuilding stadiums (McClintock, A. 2018), there are plenty of complaints related to facilities.

2. The different survey results depending on the two separate types of data:
Likert scale and open-ended questions:
Open-ended questions offer little or no structure for respondents’ answers. In general, researchers
ask a question and the respondent writes down his or her answer in a box. These questions are
flexible and allow explanation although respondents may often be unwilling to give such detailed
information. On the one hand, the Likert scale lists all categories and respondents choose the
degree to which they agree. As the given data, there are 4-point scales that make respondents
feel more convenient for taking part in the survey when provided free-choice scales. As a
consequence, the results given out from two types of data are different. Specifically, the open-
ended question could be more detailed since the respondents explained their accurate feeling
through short answers while the Likert scale examines the distinct perspectives more positively.
For instance, under the given survey, the result from the Likert scale could be more precise due to
the total of 200 opinions meanwhile the open-ended questions just received solely 49 fans.
Nevertheless, the AFL Commission decided to combine these two types of data because not only
to investigate the reasons why customers dissatisfied but also to collect the outcomes thoroughly.
This combination can generate better results for organizers to understand and enhance the
problems they are facing.

3. Could the survey have been more useful? The further analyses can be done
using collected data and the next step is?
Using a survey enables the AFL Commission to design questions and receive replies without actually
communicating to every respondent. Additionally, this is a flexible and convenient tool with a structured
format that respondents do not find it difficult to answer, as well as a fair chance of getting a true reply
(Walliman, 2011). Thereby, the survey has been more useful for the organizer.

Besides, the organizer can use control charts for attributes as the further analyses which based on
collected data because the majority of problems organizer deals with relates to common causes such as
facilities and service served in stadium. This calculation is used for processes. The AFL Grand Final is a
process because it generated service for their fans. In addition, the survey characteristic is counted rather
than measured and the observations can be classified as good or bad. Hence, the organizer can use a p-
chart instead of a c-chart since there does not exist any occurrences per unit of measure to analyze the
roots and the problems.

II. Case study 2:


Concerning the given data of two Notex Manufacturing products, using the naive method, exponential
smoothing, and weight moving average is ineffective. To be more specific, toward naive method, although
this approach seems to be the most simple, the quickest, and the easiest, its result is extremely inaccurate
because it just traces the actual data and does not smooth at all (Stevenson 2018). With regard to
exponential smoothing, I could not use it because Notex Manufacturing has not used any forecasting to
prognosticate the demand for batteries and this approach relies both on the previous forecast and the
difference percentage between it and the actual series value at that point (Stevenson 2018). Nevertheless,
there is an assumption that the previous forecast in both two products is taken from the oldest value in the
statistic or the data in January. Moreover, the chosen alpha is .60, .09, and.30, respectively in order to
testify the error in MAD calculation. To conclude, the bigger alpha, the smaller error but still bigger than
Moving Average error. Hence, Exponential Smoothing is not chosen to be the appropriate forecasting
method.

Lastly, since the weight moving average is used for smoothing a lag in moving average, it seems to be
useless to anticipate the demand for those two most crucial products. In the calculation below, the weight
moving average is assumed that the third value is assigned a weight of .60, the second value weight of
0.30, and the first one weight of .10.
Toward product 1, choosing the Moving Average forecast is highly recommended due to the usage of the
most frequent data values in generating anticipation (Stevenson 2018).

As illustrated in the excel, Moving Average error witnesses the lowest error, compared with other forecast
methods which assume n=16 since n=15, the error starts to decrease significantly.
Simple Linear Regression of Product 1
80

70

60
Number of Products

f(x) = 1.34 x + 37.86


50

40

30

20

10

0
0 2 4 6 8 10 12 14 16 18 20
Month

With the trend error, or known as simple linear regression, as illustrated in the chart above, the trend
fluctuates between October and April which means this trend does not find it difficult to change. Therefore,
Notex Manufacturing needs a more responsive method and Moving Average is the most appropriate with
this requirement. Furthermore, according to William Stevenson (2018), if responsiveness is important, a
moving average with pretty few data points should be applied.

With regard to product 2, although Moving Average error with n=2 occupies the lowest element in the
shown excel chart, there is an abnormal point is compared with the remaining months, I suggest to apply
forecast based on the trend method or simple linear regression for the more precise outcome because the
simple average is affected by outliners and all value in the average are weighted equally (Stevenson 2018).
In other words, in September, Notex Manufacturing reached a peak in demand as an unusual order,
accounted for 81 chosen batteries. Specifically, the assumption of n=7 produces a much bigger error than
n=9 due to that strange data. Consequently, the Moving Average is not a trustworthy method to apply.
Espousing the chart below, there is a slight variation between the trend line and the actual data that display
an appropriate application regarding the forecasting method. However, there is an outliner that belongs to
the special high demand for batteries in September. This abnormal change has an impact on not only the
trend error but also the moving average error. Apart from this alteration or special cause, the trend error
method is greatly advised.

Simple Linear Regression of Product 2


90
80
Number of Products

70
60
50
40 f(x) = 0.35 x + 36.67
30
20
10
0
0 2 4 6 8 10 12 14 16 18 20
Month
III. Reference:

- Goodman, T. (Ed.) 2007, The Forbes book of business quotations, Black Dog & Leventhal
Publishers, New York.
- McClintock, A 2018, ‘Stadium wars: the comedy of errors behind Sydney’s shiny new plans’,
TheGuardian, 11 April, viewed 28 March 2020,
<https://www.theguardian.com/cities/2018/apr/11/stadium-wars-comedy-errors-behind-sydney-new-
plans-football-nrl>.
- Sarkar, A., Mukhopadhyay, A.R & Ghosh, S.K 2013, ‘Issues in Pareto analysis and their resolution’,
Total Quality Management & Business Excellence, vol.24, no.5-6, pp. 641-651.
- Slameto, S 2016, ‘The Application of Fishbone Diagram Analisis to Improve School Quality’,
Dinamika Imu, vol. 16, no. 1, pp. 59-74.
- Stevenson, W.J 2018, Operations management, 13th edn, McGraw-Hill Education, New York.
- Walliman, N 2011, Research Methods: The basics, Routledge, Taylor & Francis e-Library.

You might also like