IS428 AY2019-20T2 Assign KHEMKA SHIVIKA

LIBRARY SURVEY 2018

PROBLEM & MOTIVATION

PROBLEM

The SMU Libraries are used for several purposes ranging from information access and a quiet place to study to computer and printing services. The library conducts a survey every two years to try and enhance their service quality. They ask some pertinent questions and collect useful data, but the reports generated are long and difficult to comprehend because of which the Libraries aren't able to make relevant improvements.

MOTIVATION

As an undergraduate student at SMU, the library is an essential part of my university experience and I would like to help the administration better understand the results of their survey. As a respondent myself, I have been given the rare opportunity to ensure that my feedback is well analysed and I would like to make the most of it.

DATA PREPARATION

MODIFICATIONS

Tool	Tasks and Modifications
Microsoft Excel	Understanding the dataset provided by the survey. Adding a Category column to combine the several different positions (For example: Undergraduate Year 1, Faculty-Professor) into the 4 relevant categories which are: Undergraduate, Graduate, Faculty, Staff.
Tableau Prep	Pivoting the dataset so as to bring the different rows for each question to columns. This make analysis on Tableau easier. Renaming of fields to make them more intuitive. For example: Campus to Library, ID to Exchange Student. Removal of seemingly unnecessary fields. For example: NA1, NA2. Joining the legend sheet with the data sheet so that each question text can be available and we do not have to rely on its ID. Grouping questions based on Service Categories used in the Be Heard Survey Report. For example: Questions 1,2 and 3 under Communication.

TABLEAU PREP FLOW

DATA TRANSFORMATION

Original Data	Processed Data
Size: 2639 rows and 90 columns	Size: 150423 rows and 12 columns

STORYBOARD

SKETCH 1: Survey Overview Dashboard

SKETCH 2: Survey Results Dashboard

KEY TECHNICAL CHALLENGES & MITIGATION

No.	Challenge	Description	Mitigation
1.	Lack of Familiarity with Tools	Everyone in the group do not know how to program in RShiny for visualisation	We will learn Rshiny during class, call for consultation and rely on Googling for any programming challenges. Alternatively, there is also Datacamp available for us.
2.	Viability of Ideas	We do not know if the current dataset is sufficient in providing all the information needed to conduct analysis and building of planned visualizations.	There are multiple dataset online to use and we can use Prof Kam's REALIS dataset provided to us to supplement our dataset if we are lacking of certain variables. We could also derive our own variables based on the current dataset if needed (e.g. Geocoding).
3.	Lack of Domain Knowledge	HDB resale prices are affected by a spectrum of different factors such as policy measures and redevelopment. It is hard for us to understand without domain knowledge.	Learn from informative websites such as from HDB and iteratively discover and learn insights into the dataset

COMMENTS

Do leave a comment on how I can improve my visualizations so as to be able to provide more value to the SMU Library Administration or if you require any files for reference!

No.	Name	Date	Comments
1.	(Name)	(Date)	(Comment)
2.	(Name)	(Date)	(Comment)
3.	(Name)	(Date)	(Comment)