Department of Computer Science
Vancouver BC
V6T 1Z4
|
The University of British Columbia Department of Computer Science |
201 - 2366 Main Mall Vancouver BC V6T 1Z4 |
Principal Investigator: Dr. Gail Murphy, Dept of Computer Science (murphy@cs.ubc.ca, +1 604 822 5169).
Co-Investigator: Mr. Nicholas Sawadsky, Dept. of Computer Science (nicks@cs.ubc.ca, +1 604 618 2021). This research is part of Mr. Sawadsky's research towards an M.Sc. degree. The results of this research will be included in Mr. Sawadsky's M.Sc. thesis, which will be made available to the public in due course.
Reverb automatically recommends previously-visited web links to developers as they work within the IDE. Reverb uses the frequency and recency of web page visits to generate recommendations, as well as the estimated relevance of a web page to the code the developer is currently viewing. The goal of this study is to evaluate the effectiveness of Reverb in supporting software developers as they work.
For the study, you will be asked to download and install Reverb on your computer. The tool consists of an extension to your internet browser, a plugin for the Eclipse IDE, and a separate indexing service.
During installation, you will be prompted whether you would like Reverb to index your browsing history from the past three months. Choosing "Yes" will improve the quality of Reverb's initial recommendations dramatically.
As you visit web pages, the browser extension will send the page content to the indexing service running on your computer, and the page will be indexed in the background. As you work on Java code inside Eclipse, the Eclipse plugin will recommend pages you have previously visited which may be relevant to your current code.
The tool creates logs of events occurring in the browser and in Eclipse. The logs contain web locations visited in the browser, the recommendations the tool generated, and whether you clicked on a given recommendation. Recommendations are generated and logged even if the Reverb view is not currently open in Eclipse. The logs are fully anonymized: no code or web page details (e.g. page content, URL) can be recovered from them. Web pages visited and links recommended are encoded as integers.
Periodically, the tool will ask your permission to upload the anonymized logs to a server. You will have the chance to review the contents of these logs prior to upload.
At any time, you can terminate your participation in the study by uninstalling the tool. After 6 hours of active coding have elapsed, you will be invited to provide feedback on your experience using the tool. When providing feedback, you will be asked to rate the usefulness of any recommendations you clicked on. You will also be presented with a survey form, which contains questions on your experience using the tool.
Beyond your usual coding activities, the time required to install the tool, explore interesting recommendations, and complete the feedback form will total about 30 minutes.
The only known risk is the loss of time required to install the tool and complete the feedback forms.
No information which might allow a specific web page to be identified is included in the logs generated by the tool. These logs will be kept confidential. You will be identified by numbers in any written publications. All logs collected will be kept in locked rooms, locked filing cabinets, or access-protected by password.
If you have any questions or desire further information with respect to the study, you may contact Dr. Gail Murphy at +1 604 822 5169.
If you have concerns about your treatment or rights as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at +1 604 822 8598.
By selecting "I agree to participate", you confirm that you:
| Version: March 1, 2012 |