Cryptography Reference
In-Depth Information
3.
Perceived Necessity:
did users deem the password manager necessary and accept-
able? (2 questions were posed for all PMs. For Phone and USB, 1 additional ques-
tion was posed regarding the accessibility to mobile devices.)
4.
Perceived Security:
did users find the password manager secure? (4 questions were
posed)
The users were also posed a few open-ended questions, in each of the above ques-
tionnaires, in order to poll for their opinions about any perceived problems with the
password managers and suggestions for possible improvement.
Finally, a
Final Test
questionnaire was also presented to each user polling which
password manger he/she preferred the most and asking about their order of preference
based on the level of (1) security, (2) convenience and (3) overall experience.
The main challenge we faced was the sheer number of questions which each user
needed to answer, potentially leading to lazy respondent behavior and user fatigue. Care
was taken so as to minimise both the number of questions and to discard any questions
which showed a tendency of not being answered genuinely (or honestly).
4.1
Study Participants
We recruited a total of 20 participants: 10 technical and 10 non-technical users. In the
rest of this paper, we will refer to our technical users as Students, and non-technical
users as Non-Students, because all technical people were students while all non-technical
people were non-students. The participants were recruited on a first-come first-serve
basis from respondents to emails. Prior to recruitment, each participant was briefed on
the estimated amount of time required to complete the tests and on the importance of
completing the tests in its entirety.
The student participants were all university students, studying towards undergrad-
uate, Master's and Ph.D. degrees in Computer Science or closely related fields. This
group of our users represented a fairly young, well educated and technology-savvy
sample of user population.
The other group, consisting of the non-students, had an average age difference of
nearly two decades from that of the students. This group was tested to gain insights into
whether such a group - differing in terms of full time occupation - had any impact on
the choices made with respect to the password managers.
There is an obvious concern that, if a technology-savvy group (students) does not
react well to a password management approach, the approach will perform a lot worse
with average users; or on the contrary, if a password manager that fares well with stu-
dents, it might not perform equally well with average users. This concern was our prime
motivation to categorize the respondents into students and non-students.
Our non-students ranged from help-desk personnel, technicians, real estate agents,
restaurant workers to housewives. In addition to the students vs. non-students distinc-
tion, our sample was also controlled, as much as possible, in terms of other important
user-centric characteristics, i.e., gender and age. This was done in order to evaluate the
password managers among a diverse user population pool. The gender split was: 60%
male and 40% female for both students and non-students. Our test users were divided
into three age groups: 40% Young (less than 24 years old), 40% Middle-Aged (25-44
Search WWH ::
Custom Search