ISSS608 2016-17 T1 Assign2 Ong Han Ying - Conclusion

ISSS608 2016-17 Assignment 2 - Ong Han Ying

About

Executive Summary

Introduction

Data Preparation

Approach

Results

Conclusion

Section 9:Future Work

This analytical work can be further improved, with the following potential future work;

1. Improvement in Dataset

• It was observed that the data in the research paper differ slightly as those descriptions on the dataset's website. Since the research paper is based on the data, it is essential to ensure that both are aligned.
• Sample size: Differs significantly among the different domain, and therefore; making it quite difficult to conclude and compare across the different domain. Therefore, more survey can be conducted with people from the less dominated domains.
• No. of University: There are only 2 participating university. Therefore,to have a less bias result, a similar survey should be conducted with other universities.
• Position: Only participants from UOC has filled up this field. Therefore, this field should be enforced to be completed by all participants, so that we will not have too much "unknown" which are difficult to design actionable based on it.

2. Improvement in Technical & Software Selection

• Many of the selection of graphic and interactivity are restricted to the limitation of Tableau, as it is the software that

• is able to produce most of the visual graphs
• able to create a dashboard and storyboard easily
• ability to upload onto the Cloud

• Many of the software (such as Parallel Set & Macrofocus) are good on its own, but unable to integrate with ready software. Thus, while they have great features, however; we are unable to integrate them into our dashboard
• Similar software such as Power BI and QlikSense should be explored and compare with that of Tableau
• Alternatively, we can consider self-designed dashboard using open-source programming such as d3.js

• A good example can be seen from fellow coursemate's work - Nguyen Tien Duong's Work and Nguyen Tien Duong's Assignment Write-up

Section 10 : Conclusion

In all, we can conclude that:

1. Multiple Stakeholders can read the same set of data-source differently;

• Due to differences in agenda & motivation. For instance, while "performance" is an important indicator for all the 3 stakeholders, it drives different decisions for the different stakeholders.
• Therefore, it is very important to consider the motivation of the audiences carefully.
• Also, it is important to identify different stakeholders that have an interest in the analysis, and do not expect a single dashboard to be used by all (which can lead to "over-crowded" dashboard & not being user-friendly).

2. Visual Analytics provide more insight to the story, as compared to Statistics Analytics;

• As proved in objective 2 - our analysis has shown that there are exceptions to the hypothesis (that has been accepted) by Statistics Analytics.
• Therefore, it is important for the stakeholders to realize that the research paper may not be 100% conclusive in identifying the factors, and there are exceptions to it. Hence, stakeholders have to consider this point when they are making decisions based on statistics analysis.
• It will be recommended to the stakeholders to consider the finding and zoom into the details (Similar to how we have do so in Objective 2 of this assignment), instead of basing entirely on statistics analysis.

3. Designing and Implementing Visual Analytics:

• A visual analytics is more than putting selections of beautiful graphics into a frame. There are many decisions to made, as there are many elements that can contribute to the overall visual impact of the graphic. A small element such as the graph's label or orders, has an impact on the readibility of the graphic.
• As proved in this assignment, it is not a sequential process. Thus, re-iteration is expected & user feedbacks (other than the authors) is important in improving the graphics.
• Identifying the business needs is essential, so as to ensure that the analysis is able to provide insights to the stakeholders.
• Last but not least, simple and clean layout is expected because most of the readers may not have experiment with analytics - and thus; for them to find the dashboards (and storyboard) to be useful - the dashboards must be clean and simple for them to use.

Section 11 : Reference

11.1 : Visual and Graphs

1. Divergent Graph: [[1] retrieved on 25 Sep 16.
2. Best practice to Present Market Survey Results [[2] retrieved on 25 Sep 16.

11.2 : HTML Tips and Help

1. Help : Table [[3]] retrieved on 7 Oct 16.
2. Extension : No Title [[4]] retrieved on 7 Oct 16.
3. HTML Color Code [[5]] retrieved on 24 Sep 16.
4. Help:Wiki markup [[6]] retrieved on 7 Oct 16.

11.3 Website Design

1. 10 Beautiful Website Color Palettes That Increase Engagement [[7]] retrieved on 8 Oct 16.
2. Creative Color Scheme [[8]] retrieved on 7 Oct 16.
3. Top Font for Website Design [[9]]retrieved on 7 Oct 16.

Section 12 : Acknowledgement

Special thanks to:

• (Project Mates) Mr Chia Yong Jian & Mr Nguyen Tien Duong in all the technical bits of help and tips. All these little help matters to a technical noob.

• Course Mates :

1. Ms Lee Gwo Mey - the author has made reference to her work, and improved on

• the current version of the divergent chart, "line" that cut between the "neutral" responses.
• present a shorter version of the questions on the dashboard, instead of the full length of the questions in the survey.
• Click HERE to view her great work!

---

Previous Section: Results