Information Technology Reference
In-Depth Information
This low value of alpha led us to remove an item pertaining to writing down one's password, and
this increased that latter alpha reliability estimate to .68. We analyzed the password-writing item
separately.
Next we conducted analyses to explore the effects of organizational circumstances and per-
sonal demographics on the behaviors. In the first analysis we used MANOVA to examine effects
of geographic location, size of company, and type of industry on password management, password
writing, password sharing, and organizational support for security. Multivariate results indicated
that company size, Wilk's lambda
.88, F(40,1894)
1.5, p
.05, and company type, Wilk's
lambda
.05, had statistically significant effects on the outcomes. In
overview, results showed that employees in larger organizations reported better password man-
agement practices (more frequent password changing and selection of stronger passwords); employ-
ees from the military, financial institutions, and telecommunications/Internet companies reported
better password management practices than employees in other organization types; and more
security support was perceived in the military and public utilities than in other organizations.
Finally we examined demographic factors: Multivariate tests revealed that job type, job tenure,
union membership, age, and income level had statistically significant effects on the outcome vari-
ables, with Wilk's lambda values ranging from .95 ( job type) to .99 (union membership). Admin-
istrative personnel, managers, and technicians reported better password management than people in
other job types, but administrative support personnel, managers, first-level supervisors, and sales/
management personnel also tended to share their passwords with others more frequently than those
in other job types. Individuals with longer service reported greater organizational support for secure
behaviors (e.g., availability of training). Union members reported slightly less favorable password
management practices than those not in unions. Those with higher incomes reported better password
management practices and less password sharing than those with lower incomes.
The MANOVA and regression findings described above suggest that organizational and demo-
graphic factors relate to security behavior. Individuals in organizations that had “a lot at stake”
(e.g., in military and financial institutions) had apparently received the training, tools, and incen-
tives needed to enact more effective information security behavior. Likewise, individuals further
up the organizational hierarchy may have both a greater personal investment in the organization
as well as access to more resources (such as training) than those further down or with a shorter
history in the organization.
.88, F(56,1943)
1.8, p
Survey Study 2
As in the first survey study, we focused on asking about novice behaviors (e.g., choosing a hard-
to-guess password). In addition to asking about password management, the eight security-related
behaviors on this survey included several behaviors not assessed in survey study 1: three types of
counterproductive computer usage, discussing security policies with coworkers, and abiding by
acceptable use policies. The survey randomly sampled eight hundred employed adult respondents
from the StudyResponse online panelist project (http://www.StudyResponse.org). We intentionally
sampled individuals who reported their jobs as not in the military, utilities, financial institutions,
or telecom so that we would not need to control for organization type in this research. Given the 298
usable responses we obtained, this procedure yielded a response rate of 37.25 percent. Participants
completed a brief survey with measures of organizational commitment (Allen and Meyer, 1990),
assessments of positive and negative feelings at work (Miles, Borman, Spector, and Fox, 2002), a
self-assessment of technical know-how, demographic data, and a sample of eight of the novice
behaviors from our security behavior list.
Search WWH ::
Custom Search