Proposal
Title
Widget Usability
Lay Summary
Widgets are important components of modern software tools. For those tools to work effectively, widgets must be easy to use. This project gathers data from users and computer professionals to investigate how usable current Widgets are and to provide evidence to inform how to make future Widgets more usable.
Background
Widgets have been used for many years (Andrews, 2012), and include three main types (Abelian, General Universal and Zoomorphic) (Bloggs and Cameron, 2011). However it has been recognised (Dunce and Egghead, 2009; Flycatcher et al., 2010) that Widgets suffer from usability problems which hamper their re-use and constrain their development. Although much research about Widgets is ongoing, including work at Erewhon University (erewhon.ac.uk/research/widgets) and Lilliput Institute (lilliput.ac.uk/groups/wrg), usability has not been addressed.
Aims and Objectives
This research aims to develop a novel framework for Widget usability and to implement and evaluate an architecture to support the framework. Objectives include
- Data collection from users and developers to identify the major problems with Widget usability;
- Creation of a framework informed by analysis of the data;
- Creation of an architecture to support the three major types of Widget within the framework;
- Implementation of the architecture using current technologies, including Java, EJB, Apache and Python;
- Evaluation of the architecture.
Design/Methodology
There are two section of this project which involve human participants - the initial data gathering and the final evaluation.
For the initial data, we envisage delivering a questionnaire to 100 final year students at Warwick University and interviewing 10. The questionnaire will generate both quantitative and qualitative data, the interviews qualitative data. These data sets should be sufficiently large to gather sufficient data, and are consistent with usability research for other types of software, such as employed by Giggle and Laughalot (2003).
Statistical correlation tests will be used on the quantitative data to test the hypotheses we will develop as to Widget usability, and content analysis will be performed on the qualitative data to identify themes affecting Widget usability.
The questionnaires will be distributed at core lectures at the start of term 2 this year. Respondents will be clearly informed that they are not required to complete the questionnaire, and if they do it will be anonymised. A PIL is attached which details the process for the respondents.
Volunteers will be sought to be interviewed, and will be informed that their data will be stored fully anonymised, and that if later they desire to withdraw from the study, any data collected from them will be destroyed.
Ethical Considerations
Given the nature of this research, good practice for delivering questionnaires and interviews will be followed, as detailed in Cohen and Manion (2012).
- Informed consent. Participants will be informed clearly what the purpose of the research is via the PIL, and for interviewees a consent form will additionally be completed.
- Participant confidentiality and data security. Data on all participants will be stored anonymised, paper copies of data in a locked filing cabinet in room CS399 in Computer Science (accessible only by the researcher and their supervisor, Dr C Clever), and electronic data on an encrypted filesystem on the Computer Science system, for which the encryption/decryption keys are held only by the researcher and their supervisor. Anonymisation will be performed via a one-way hash function on the student's University ID number. Data will be stored for 10 years, as required by BSREC.
- Right of withdrawal. Interviewees will be allowed to withdraw (and their data destroyed) at any time after the interview and prior to the results of the research being published.
- Process for dealing with sensitive disclosures. Not applicable - the data are not sensitive.
- Benefits and risks. Participants will be invited to receive a copy of the final report / papers. There are no known risks associated with participating in this study.
- Further issues. Not applicable.