This was a new question for this survey. Respondents were asked to indicate the primary place
from which they access the WWW. Only one answer could be selected. "Distributed" means that they
do not have a primary place--their access is distributed.
More than half of the respondents said their primary place of access was at home (55.4%).
This number corresponds well with similar questions, such as Who Pays for
Your Access where over half of respondents indicate that they pay for their own access.
More European users than US users consider work to be their primary place of access (46.3% vs.
Slightly more females have some location other than work or home as their primary place of
Web access (11.7%).
More than 73% of those over 50 reported that their primary place of access is at home.
Respondents in the 26-50 age group are more likely to have their primary access from work
than any other age group (39.2%).
A larger percentage of younger respondents have some other source as their primary access (20.3%)
compared to other age groups (approximately 2%).
This question attempts to try to determine the primary provider of Internet access to
the respondents. Major online providers are Prodigy, Compuserve, etc., whereas local
providers usually only provide connectivity to a regional area. "Commercial" refers
to gaining access through primarily through work rather than one of the other
We continue to see growth in the number of respondents using local providers (48.5% fifth, 41.6% fourth). The next largest categories are educational providers (26.8%) and major providers (9.24%). The percentage gaining access from work dropped several percentage
points since the fourth survey (7.9% fifth, 10.0% fourth).
European respondents reported virtually the same distribution of sources as in the
fourth survey. In the fifth survey, a smaller percentage of US respondents reported educational providers a primary source (26.7% fifth, 32.5% fourth).
More males than females report gaining access through local providers (50.9% male,
43.4% female) and vice versa for educational affiliations (31.4% female, 24.7% male).
These differences are less pronounced than in the fourth survey.
More than half of those aged 19-25 reported educational institutions as their
primary Internet provider, but almost 30% report using local providers as well.
The majority of users over 50 cite local service providers as their primary
Older users are somewhat more likely to subscribe to a major Internet service
The average age of all the Web users that responded to the Fifth
Survey is 33.0 years old, a slight increase from the Fourth Survey,
which had an average age of 32.7 years old.
One of the re-occurring differences between European and US users
is age profiles, with the European users (average age 28.8) being
younger than their US counterparts (average age 33.9). These averages
are almost the same as observed in the Fourth Survey (European 29.7; US
As with the Fourth Survey, there are few major differences
between the age profiles for man and women. The average age for women
is 31.9 years old, which is slightly younger than the men, which have
an average age of 33.4 years old. These numbers are almost identical
to the Fourth survey where the average ages were 31.8 yrs old for women
and 33.1 yrs old for men.
A trend observed in the Fourth Survey that continues in the Fifth
Survey is for the women to have a stronger presence within the 11-20
year old range, with 12.9% of the women belonging to the 16-20 yr old
category compared to 10.1% of the men. The stronger presence of younger women is supported by the occupational data which also shows a strong presence of women college students.
The below table looks at all countries/states weighted equally.
From this we see that 11.0% of the respondents came from California,
followed by Texas (4.8%), New York (4.4%) and the United Kingdom (3.2%).
This is the same ranking of the top three as in the Fourth Survey.
Compared to the Third Survey, California is exerting less dominance over
the location of the population. This may very well be attributable to
the wider acceptance of the Web outside of Silicon Valley, the heart of
the computer industry and Internet Companies.
The majority of the users surveyed report having no dependents
(61.7%). European users (74.0%) are more likely to have no dependents
than US users (61.0%). This trend was also observed in the Fourth
Another trend extending back to the Third Survey, is that more
users report having two dependents (16.4%) than one dependent (14.1%).
total of 7.9% of the users have three or more dependents.
Females reported having slightly more dependents than their male
counterparts, though this difference is not statistically significant.
As one would expect, the youngest and the eldest age groups report
having fewer dependents, with the 26-50 age group having the most. 74.6%
of the 19-25 yr olds and 73.2% of the 50+ group report having no dependents, with 60.6% for the 26-50 yr olds.
Overall, the distribution of educational attainment has stabilized
from the trend of increasingly lower levels of attainment as seen in
the Second, Third, and Fourth Surveys. The profile observed in the
Fifth Survey closely resembles the Fourth survey, with 56.5% of the
users having completed college or a more advanced degree (compared to
55.0% in the Fourth and 73% in the Second Survey).
European users (23.1% Masters, 5.4% Doctoral) tend to have more
advanced degrees than the US users (16.2% Masters, 3.4% Doctoral).
In the US, the percent of users who have completed only grammar
school increased from 1.9% in the Fourth Survey to 2.7% in the Fifth
Survey, indicating more younger users.
The differences between the educational attainment of female and
male users continues to decrease. Nearly the same number of women
and men have graduated from college (33.1% female vs 32.9 male),
with more women how have completed some college (29.9%) than men
(26.1%). However, men have more doctoral and professional degrees
(7.3% men vs 4.9% female).
Not surprisingly, the 19-25 age group reports having significantly
lower levels of educational attainment than the other groups. Likewise,
the 50+ age group (34.1% Masters/Doctoral) has more advanced degrees
than the 26-50 age group (25.4% Masters/Doctoral).
This question attempts to begin to understand the response rates and various
populations on the Internet and Web. This is the third time we asked the users
to inform us how they found out about the current surveys. We broke the data
into groups whom responded via announcements to newsgroups, other WWW
pages, and listservs, etc. to see if these groups of users are different from one
another. This may very well be one of the most interesting data points for
understanding the use of the Web for surveying. Users were allowed to select more
than one source, so the numbers may add to more than 100%.
As in the previous surveys, the largest number of respondents found out about the survey
by following a link from another Web page (69.0%). There was a significant decrease in the number who found
out through Usenet newsgroups ("usenet" 15.7% in the fourth, "nntp" 7.0% in the fifth) and a small rise
in several other categories: from friends, magazines, email, remembered from last time, and "other".
This is encouraging news for us, because a diversification of the ways people heard about the survey
suggests that we are reaching a broader audience of Web users.
More than three-quarters of our female respondents (75.2%) heard about the survey from other
Web pages compared to 65.9% of males. This is a reverse from the fourth survey where more males than
females came to the survey from other Web pages.
There were only small differences between age groups for this question.
Those in the 26-50 age range were somewhat less likely to hear about the survey from other Web pages
compared to other age groups, and slightly more likely to hear about it from friends and email.
Nearly 42% of respondents spend under 5 hours per week doing
fun computing; 27.6% spend for 6-10 hours per week and 28.9% spend 11 to 50
hours per week. Although this distribution has not changed dramatically since
the fourth survey, there is slight trend toward respondents spending less time having fun
with their computers.
Approximately half of European users (49.7%) claim to use their computers for fun
for 1-5 hours per week, compared 35.8% of US users. Conversely, twice as many US
users claim to use their computers for fun for 21-30 hours per week (8.1% US, 3.5% Europe).
For casual computer fun (i.e. less than 20 hours per week), females seem to spend
less time than their male counterparts. For intense computer fun (i.e. more than
20 hours per week) there are no differences between the amount of time spent by males
Both the oldest users and the youngest users spend more time on fun computing
than those in the middle age range.
Overall, 31.5% of the users are female and the other 68.5% are male.
This represents a moderate increase in female users from the Fourth
Survey, where 29.3% reported being female, and quite a significant
jump from the Third Survey (April 1995), where 15.5% reported being
The US segment continues to be integrating more female users into
its user base than other countries, with 34.4% of the users being female
in the US (65.6% male). Europe reports only 15.2% females. However,
this is a 45% increase from the the Fourth Survey, where 10.5% of the
European users were female.
Compared to random sampling surveys, like O'Rielly, FIND/SVP, and
Nielsen, the gender ratios for the Fifth survey are all with the
reported margin of error, i.e., no statistical differences.
The younger age groups report more females than the older groups
(32.1% 19-25 yrs old, 30.1% 26-50 yrs old, and 24.7% over 50 yrs old).
Since the Web does not easily facilitate access by users with
disabilities, it is not surprising that 91.7% the users do not report
any disabilities. This is almost exactly the same percent reported in the Fourth Surveys (91.8%).
Impaired vision is the most widely reported disability, but still
only accounted for 3.7% of the responses. The other disabilities
all were reported by less than 1% of the respondents apiece.
Compared to the other surveys, the percent of respondents without
impairments has remained quite stable. For the Third Survey, 91.8%
reported no disabilities and for the Second Survey, 95% reported no
The proportions with impairments are consistent between the
stratified segments of location and gender, but the age segmentation
did reveal some differences. Elder users reported significantly more
hearing, motor, and multiple impairments compared to the younger groups.
The estimated average household income for the Fifth Survey is $59,000
US dollars. As with the Third and Fourth Surveys, this questions received
the most 'Rather not Say!' responses (14.0%), nearly seven times greater
than any other question.
The average income for the Fifth Survey is slightly lower than the
Fourth Survey ($63,000) and much lower than the Third Survey ($69,000).
We do note that for the Fifth Survey, we changed the ranges provided to
the users of values to more accurately reflect normal income levels.
As with previous surveys, the European users have more users in
the lower income brackets (23.0% under $20k) and fewer above $50k (31.3%).
This trend is due to the strong presence of students in the European
Web user community.
Female users typically report lower income levels than their male
counterparts, especially in the over $50K income brackets.
As with the Fourth Survey, more women chose the 'Rather not Say!'
than men (16.7% female vs 12.7% male), though this was less than reported
in the Fourth Survey( 18.3%).
As one would expect is directly proportional to age, with the 19-25
yr old age group reports having less income than the older age groups
(29.4% under $20k). For the 26-50 yr old age group, 48.4% report a
household income above $50k, with 61.5% over $50k for the over 50 yr
old age group.
Overall, 43.1% of respondents have been on the Internet for less than one year,
which is down somewhat from the fourth survey, where 60.3% had been on for less than
a year. Close to a quarter (23.6%) have been on for less than six months.
34.9% of respondents report having been on the Internet for 1 to 3 years -- the
highest percentage ever reported in this range. This suggests that the flood of new
Internet users seen over the past year is slowing somewhat.
The continuing influx of females to the user population is illustrated by the
fact that 56.6% of female respondents report being on the Internet for less than
one year, and 34.0% for less than six months.
Male are still more than twice as likely as females to have been using the
Internet for more than 4 years (26.5% male, 12.1% female).
Another group of users that is growing is those over age 50 with 34.1% having been
on the Internet for less than 6 months.
The largest group of 19-25 year olds have been online for 1 to 3 years (41.1%).
Overall, 88.6% of the users report English being their native/first
language. For the US, this number increases to 96.1% and for Europe it
decreases to 43.3%. Europeans reported the 'Other Language' category
47.9%, indicating a problem with our question and response choices.
Given the tendency of female users to be located in the US, it is
not surprising to see an increase the in the number of native/first English
speaking females (94.1%) in the sample compared to men (85.6%).
The same location bias that occurred with female users appears to also hold
for the 50+ yr old age group, with 94.6% native/first English speaking compared
to 82.6% for the 19-25 yr old age group.
For classification of location by major geographical location,
73.4% of the respondents were from the US, 10.8% from Europe and
8.4% from Canada & Mexico. Compared to the Third Survey which was run
one year ago (80.6% from the US, 9.8% from Europe, and 5.8% from
Canada & Mexico), this represents a significant shift towards less
of a US dominance in Web users.
Additionally, notable increases occurred in most of the other
geographical areas like Asia, Africa, Oceania, the Middle East, and
Central & South America. Responses were received from all the
As evident from the Gender analysis, the US has percentage-wise
more female users than all other locations. Over 80% of the female
users were from the US, with 70.3 of the male users being form the
The vast majority of elderly users are located in the US (83.4%),
more so than the other age groups (74.9% 26-50 yr old and 67.1% 19-25
yr old). Besides the US, the largest concentration of younger users
(19-25 yr old) is in Europe (16.6%).
One of the more stable characteristics of Web users over the
survey is marital status. Overall, 41.1% of the users are married,
with 40.8% being single. The users whom reported living with another
was 9.6% and those reporting begin divorced was 5.1%. Europeans were
twice as likely to report living with another person (18.4%) compared
to the US (8.0%). These percentages are almost exactly the same as in
the Fourth Survey.
As with the Third and Fourth Survey, women Web users are less
likely to be married than men (38.2% verses 42.5%) and more likely to
be divorced (7.0% verses (4.2%) or living with another person (10.8%
As one would expect, three quarters of the 19-25 yr olds are single, with
three quarters of the 50+ yr olds being married. The 26-50 yr olds are more
likely to be married than single (54.0% vs 25.6%).
More than 96% of respondents reported using a color monitor which is probably good news
for content designers. Unfortunately, 33.8% were unsure of their bit depth. Of those
who did know, 38.9% reported 24 bit color and 18.5% reported 16 bit color.
While respondents seem to have the hardware to support graphic-intensive content,
content designers need to consider the real limiting factor for most users: Speed of Connection to the Internet.
Twice as many females were unsure of their bit depth than males (54.2% female, 24.4% males).
Because so many respondents in all categories are unsure of their bit depth, the
differences in the other categories are not very significant.
The majority of respondents (53.2%) reported that they owned monitors that were
13-15 inches in size.
European users owned more larger monitors than their American counterparts. Almost
30% of Europeans use 16-18 inch monitors as opposed to only about 24% of US respondents.
Similarly, almost 12% of Europeans use 19-21 inch monitors, where just over 7% of
Americans use them.
It is unclear whether females, in general, use smaller monitors than males
because so many more females were unsure of their monitor size (10.1% female, 1.8% male).
The most common monitor size for all age groups was 13-15 inches. Younger respondents
seemed to have larger monitors, in general.
Nearly half of all respondents own only one computer (47.5%), while over
a quarter (26.9%) own two. There numbers are almost identical to the fourth
survey. Overall, there was a slight decrease in the number of respondents who
don't own any computers (10.5% fourth, 8.1% fifth).
As with the fourth survey, the differences between US and European users
in terms of PC ownership are insignificant.
Males are more likely to own multiple computers. Twice as many males own
3 or more computers (21.2%) compared to females (9.6%).
Females are more than twice as likely as males not to own a computer (13.7% female,
Respondents over 50 were more likely to own multiple computers than any other
age group. Respondents in the 19-25 age group were least likely to own any computers.
As with the Fourth Survey, Educational occupations account for 29.6% of
the Web users, with Computer related occupations a close second at 27.8%.
This is significant shift from the Third Survey a year ago where Computer
related occupations accounted for 31.4% of the users and Educational occupations
accounted for 23.7%. Thus, there appears to be a solid migration of
non-computer science users, with a strong inflow from the educational sector.
Professional and management occupations account for 18.9% and 10.7%
European users are more likely to be in an Educational occupation than
Computer related (37.0% vs 31.0%).
Female users tend to be primarily involved in Educational occupations
(35.5%), with Professional (19.8%) and Computer related (18.2%) occupations
following. This is quite a different occupational profile than males, and
is supported by the educational attainment and age profiles of females.
Male users are slightly more likely to be in Computer related (32.2%)
occupations than Educational (26.8%), with Professional (18.5%) and
Management (11.5%) following.
Overall, 51.6% of users do not subscribe to an online service. This is up slightly
from the fourth survey in which 47.7% did not subscribe. The largest categories for
users who do subscribe are: "Other" (24.7%), America Online (16.6%), and Compuserve (11.0%).
An even higher percentage of European users do not subscribe to any online service (59.2%) or to services not listed in the question (26.9%). Of those listed, the largest
categories for European users were Compuserve (14.4%) (which is actually higher than the
US number for Compuserve: 11.1%) and Microsoft Network (3.9%).
For the services listed, the largest categories for US users were: America Online (22.0%), Compuserve (11.1%), NetCom (6.5%), and Microsoft Network (4.1%).
Female respondents were slightly more likely than males to subscribe to some
Respondents over age 50 were more likely to subscribe to an online service than
other age groups. For each service listed, respondents aged 19-25 had the lowest
percentage and respondents over age 50 had the highest.
This question was only given to those who answered the survey from the US.
The largest category was "democrat" with 25.4% closely followed by "republican" with
21.1%. The next two largest categories were "independent, leaning toward democrat" with
16.4% and "independent, leaning toward republican" with 10.3%. Only 7.4% of respondents
classified themselves as strict independents. 5.8% classified themselves as libertarians.
It is interesting to note that although most people identify with one of the major
parties, most also classify themselves as "moderates" (See Political Affiliation.)
More females than males classify themselves as democrats or independents leaning
toward democrat: 50.7% female, 37.1% male.
More males than females report being libertarians: 7.4% male, 2.9% female.
Respondents over age 50, in general, identify more strongly with their party of
choice than do other age groups; 42.9% classified themselves as clearly democrat or
A higher percentage of those between 10 and 25 said they didn't know what
party they preferred (7.3%) when compared to other age groups (3.0% for 26-50, and 1.5%
for over 50).
This question has changed since the last survey, so a strict comparison of answers is probably
not fair. In previous surveys, we asked if respondents willingness to pay depended on the cost and/or
quality of the information provided. This time, we presented several different payment schemes to find
out what schemes users preferred.
With each survey, the percentage of respondents who have stated outright that they would not
pay for access to WWW pages has been increasing. For the fifth survey, 65% said that they would not
pay. This may reflect the fact that most people primarily use the Web as a source of entertainment and
not necessarily a resource they are willing to pay for. Another reason might be the fact that so many users
are now paying Internet service providers for Web access. They may not be willing to pay twice: once for
access to the Web in general and again for specific Web pages.
For respondents would would agree to pay fees for Web pages, the most popular models were
a subscription model (12.1%) and pay-per-view (10.9%).
Users in Europe were more agreeable to paying for Web pages, perhaps because more of their
access is subsidized by universities and businesses.
A slightly higher percentage of females indicated that they would not pay for access to
particular WWW sites.
The youngest users were least inclined to pay for access to Web sites (69.0%).
For those willing to pay, the subscription model was favored by the youngest users (13.0%),
while older users preferred a pay-per-view model (13.7%).
For this question, respondents could choose more than one answer.
Even more respondents than last time report paying for their own Internet
access (57.7% fifth, 51.0% fourth). This is followed by having it paid for by
work (30.3%) and school (19.3%).
European users are still more likely than US users to have their access paid
for by work (43.2% Europe vs. 28.4% US), but more Europeans are paying for their
own access than in the last survey (42.6% fifth, 38.6% fourth).
More males than females are paying for their own access (61.4% males, 49.4% females).
Females have their access paid for by school, parents, or other sources. This explained
by the observation that many of the female respondents reported being students or
otherwise affiliated with education.
The majority of users over age 50 (80.7%) reported paying for their own Internet
As one might expect, a large percentage of users between ages 19 and 25 have their
access paid for by school (44.7%). A significant amount, though, (38.9%) are paying
for it themselves.
More users between ages 26 and 50 have their access paid for by work (36.8%) than
any other age group, but many pay for it themselves as well (66.1%).
Once again, more than half of all respondents (58.6%) reported using some flavor
of Windows (3.1, 95, or NT) as their primary computing platform. This percentage is
down slightly from the fourth survey (61.5%). The second most common is Macintosh
with 28.2%, up from 20.5% in the fourth survey. Unix is a distant third, with only 5.1%.
Consistent with previous surveys is the observation that significantly more
Europeans use Unix than Americans (12.0% vs. 4.4%). Also, there are still more
US Mac users than European Mac users, even though Macintosh usage gained several
percentage points in Europe (14.1% fourth, 20.2% fifth).
There are almost as many respondents still using Windows 3.1 (27.4%) as Windows
More females use Windows products than males, especially Windows 3.1, and more
males than females use Unix.
Interestingly, there are more male than female Macintosh users (30.6% vs. 23.1%).
This is a reverse from the fourth survey where only 17.7% of males used Macintoshes.
Older respondents are more likely to use a Windows product, while younger respondents are more
likely to use some version of Unix.
Overall, the largest category of respondents considered themselves moderate in
their political views (30.1%). 21.1% considered themselves to be conservative or
very conservative, while 35.18% were liberal or very liberal.
For US respondents, the curve peaked at moderate (32.5%) with 4.4% in the
conservative extreme and 8.9% in the liberal extreme. In Europe, however, the
curve peaked at liberal (33.6%) with only 0.7% in the conservative extreme, but
17.1% in the liberal extreme. One thing to consider with these numbers, though,
is that the terms "liberal" and "conservative" may have slightly different meanings
in different cultures, so a strict comparison between the US and Europe may not
More females than males reported being liberal or very liberal (40.6% female, 32.7%
male). About the same number, however, reported being moderate.
The largest category for all age ranges was moderate.
More respondents over age 50 reported being conservative or very conservative (34.0%) than any other age range.
The majority of Web users report their race as being caucasian/white (87.3%).
The other races did not show any notable increases since the Fourth Survey,
indicating that little has changed in the past six months with respect to this
characteristic. In the US, 88.6% report being caucasian/white, 2.7% asian,
and 1.3% african american/black.
There are no differences between gender with respect to race.
A definite age effect occurs with race. The 19-25 yr old are less likely
to be caucasian/white (83.4%) than the other age groups, with 5.3% reporting
being asian. The eldest group was the most likely to be caucasian/white
(95.7%). The 26-50 yr old age group typically falls between the two groups
Overall, the response profile across the categories is relatively flat. Only 25.1%
spend less than 5 hours of computing time on work, while 18.5% spend over 41 hours.
This is almost identical to the profile from the fourth survey.
As with the fourth survey, European respondents spend more hours per week working
with computers than do US respondents: 56.4% of Europeans spend more than 21 hours
per week compared to only 44.8% of Americans.
Females generally spend less time using their computers for work than do males.
Over a quarter of females spend 1-5 hours, opposed to just over 17% for males.
Twice as many males spend over 50 hours per week than females (10.1% males, 4.7% females).
Users in the 26-50 age range rely on computers for work much more than the other
age ranges. Nearly 40% of 26-50 year olds use computers for work for more than 30
hours per week.