They say we're in a revolution. The Internet and personal computers are turning futuristic fantasies into everyday realities at a dizzying pace. But where are the facts?
Solid numbers to track the growth and behavioral impacts of personal computers are hard to find. Sales data and "hits" on Web sites raise more questions than they answer. Now, two telephone surveys of individuals contacted randomly across the population have arrived to shed new light on the impact computers have on daily activities, and who chooses to spend free time with their home computer.
Market researchers make enormous efforts to find out how to get products into homes. They are not as concerned about what happens after the product is in the home. But two recent national surveys have expanded our knowledge of computer users and the activities possibly affected by the personal computer. These surveys are large enough to distinguish heavier computer users from lighter users. Moreover, they can be statistically controlled for such major demographic predictors of computer usage as gender, age, education, and income.
Another major question is whether owners of personal computers and users of online services are reducing their use of the traditional mass media, arts, sports facilities, and other free-time activities. The two surveys can identify which types of media have been affected by computers, and they can suggest some ways they have been affected by the arrival of these new technologies.
The 1995 Times-Mirror Survey and the 1997 Survey of Public Participation in the Arts (SPPA) focus on a broad range of daily activities related to computer usage. Instead of asking how their lives have changed as a result of computers, respondents are asked to simply estimate what they do. One advantage of this approach is that it covers both general use and use on specific days. Moreover, the activities covered in these surveys consume more than two-thirds of the free time that respondents report in full-time diaries. These surveys reveal whether people who use their personal computer for hobbies and enjoyment also report lower levels of other free-time activity.
When the 1997 SPPA asked Americans whether they used a computer at home during their free time, 61 percent said no. Another 18 percent of the sample played on their computers less than two hours a week; 11 percent spent 3 to 5 hours a week; and 10 percent spent 6 hours a week or more. The overall average estimated time was 2.0 hours per week for the whole sample, and 5.2 hours for those who used personal computers at all during their free time. Clearly, home computer use is an activity where actual use still falls short of marketers' hopes.
DEMOGRAPHIC DIFFERENCES Playing on the computer is no longer a man's world. The share of men who log on at home (42 percent) is not much higher than the share of women (35 percent). There are, however, dramatic differences by age group; 69 percent of 18-to-24-year-olds use computers for hobbies and enjoyment, compared with only 10 percent of those aged 65 and older. The biggest age drop is between those aged 45 to 54 and those aged 55 to 64, with only about half as many in the latter group using the personal computer.
As with gender, the differences between whites (43 percent) and blacks (38 percent) are not large. But there is a large difference between those who describe themselves as Asian (63 percent) and Hispanic (24 percent). Many of these differences can be accounted for by social class and age factors.
Access to personal computers is, of course, highly related to income. More than two-thirds (68 percent) of those earning more than $100,000 a year have used a personal computer at home, compared with less than 25 percent of those with family incomes under $30,000 per year. Much the same range of differences is found by education, with usage ranging from 62 percent among college graduates to only 12 percent among those who never finished high school.
The three factors that largely determine how much free time we have are work, marriage, and children. Longer work hours go along with many other free-time activities including higher levels of leisure computer usage: over 50 percent of those working at least 50 hours a week use a computer at home. People who work part-time (1 to 29 hours a week) are also likely to use home computers. Married people are slightly below average in computer usage, and the divorced, widowed, or separated are lower still in relation to never- married people. Adults with children at home, no matter how many, are more likely than nonparents to be users.
Several of the above differences are mostly due to the respondents' age or education. But the demographic basics don't explain it all. Statistical adjustment for marriage, presence of children, and work hours reveal the following: * Only slightly higher usage by men than women (41 percent versus 37 pe rcent).
* Reduced but still notable differences by age, from 72 percent of those aged 18 to 24 to 17 percent of those aged 65 and older.
* Usage rates by race that are much smaller after adjustment, although Asians are still far more computer active (50 percent) than Hispanics (28 percent).
* Income differences reduced much more than education differences.
* Work hour differences that largely disappear, except for the low levels among those working 40 to 49 hours.
* Difference by marriage and children that are also reduced, although having children is still associated with higher usage.
HEAVY USERS These differences in simple rates of usage conceal some important differences in heavier versus lighter users. For example, the 42 percent of men who use computers at home give weekly usage estimates that are 20 percent higher than do the 35 percent of female users. Age differences, on the other hand, become more level generally. Users older than 44 use computers more than those aged 25 to 34. Those aged 18 to 24, however, use computers more hours than average, increasing their free-time usage to nearly four hours per day overall. This is twice as high as those aged 35 and older, and half again as high as those aged 25 to 34. This may mean that overall usage for the population could increase as today's young people move into older age groups.
Some racial differences become more pronounced after adjustment for other demographic factors and for extent of usage. Black and Native American users show much heavier usage than do whites, Asians, or Hispanics, for example.
In general, home computer use is rather flat or irregular by education and by income. Usage is still higher among the college-educated and affluent, although peak usage appears to be those in medium education and income categories. And users at the extremes of work hours (0 hours and 50 or more hours) report the most time spent at home and on computers.
On the other hand, usage rates by marriage and children are rather similar across categories after making adjustments for demographic factors and usage. People without children may be less likely to use a computer for pleasure, but if they do, they use it more than parents with children at home. Age and education differences also account for most of the other initial differences by these family factors.
COMPUTERS AND OTHER PASTIMES The SPPA survey contained participation questions on more than 25 free-time activities, most of them related to arts events and arts content in the mass media. It found that Americans who use personal computers at home are also more active in virtually all free-time categories. For example, with respect to attending arts events, almost 18 percent of computer users had attended a live performance compared with 9 percent of nonusers. For attending a stage play, the contrast was 20 percent versus 12 percent. For the arts museums or galleries, it was 48 percent versus 30 percent.
The same pattern exists for serious reading. Reading a novel in the previous year described 71 percent of computer users, but only 54 percent of nonusers. For reading poetry, it was 34 percent to 25 percent. For electronic media, such as listening to jazz on the radio, classical music or records, or watching a play on TV, the differences were similar but in the same direction-21 percent versus 17 percent for jazz radio, 52 percent versus 40 percent for listening to recordings of classical music, and 12 percent versus 10 percent for watching a stage play on TV.
The familiar pattern also extends to non-arts activities. Movie attendance in the previous year was reported by 80 percent of computer users, but only 57 percent of nonusers. For attending sports events, it was 54 percent versus 35 percent. For home improvement, 71 percent versus 63 percent. The only free-time activity in the leisure list that showed no relation was gardening, with 65 percent of users and 65 percent of nonusers reporting. Nor did the "more-more" pattern extend to TV use in general, which was estimated by SPPA respondents at about 3 hours a day. Translated to weekly hours, computer users watched an estimated 17.5 hours of TV per week, compared with 21 hours for nonusers.
Two cautionary factors need to be taken into account in these calculations. First, most of the differences occur between computer users and nonusers, and they do not increase as estimated computer use increases. For attending a play, for example, the rates of attending are 12 percent among nonusers, 17 percent for the one-hour group, 24 percent for the 2-hour group, and 26 percent for the 3-to-5-hour group. It then drops to 12 percent for the 6-to-10-hour group and 19 percent for the 11-or-more-hour group. As expected, then, attendance does tend to drop as usage increases.
Second, the activity differences do not take into account common relations with background factors, like age, income, and education. Younger, more affluent, and more educated respondents do more of each of these activities. When these activity data are statistically adjusted for these common demographic correlates, the differences by computer use drop dramatically. However, they do not disappear and are not reversed. In other words, there is still the tendency for heavier computer users to be more active in other free-time activities, except for gardening and TV. Moreover, that pattern is consistent with the greater usage by those with longer estimated workweeks.
The overall usage of mass media is addressed in detail in the 1995 Times-Mirror survey. In that survey, questions on long-term computer usage are supplemented with questions on computer and Internet usage "yesterday," where those yesterdays were scattered across the week. The overall estimates for yesterday were 13 minutes of computer use and 4 minutes of Internet use. That translates to the same 2.0 hours per week as in the 1997 SPPA, even though the questions cover different computer users.
The demographic profile correlates in the Times-Mirror and SPPA surveys are similar with lower age, higher income and education, and men corresponding with higher rates of use. Differences by race were not as pronounced in Times-Mirror, although again marriage and children differences were minimal. Work hour differences again show that more active users tend to have longer work hours.
The pattern of results for mass media differences shows much the same pattern as for arts activities. As computer usage goes up, more use is also made of the news media, and particularly for print media, such as newspapers, magazines, and books. Use of the radio for news is also higher, although TV use in general is slightly negatively related.
Again, these patterns held up after statistical adjustment for the main demographic correlates. And they also held up for longer-range estimates of computer usage, that is, those who said they use computers and the Internet "every day" or almost as often.
The two studies show about the same amount of time on personal computers, although the SPPA data exclude time spent on work-type computer tasks. They also show the same set of demographic correlates, and the same "more-more" pattern of relation with other free time. It does not seem that computer use cuts into any of the other ways of spending free time measured in the survey. Computers even seem to coexist with the 800-pound gorilla of free time: television.
Where do the new hours devoted to computer use come from? The answer isn't clear. Not all free-time activities have been examined, yet computers could cut into one's social life, religion, organizational participation, or adult education activities. Also, there are nearly 100 hours a week devoted to non-free-time activities. Computer play does not cut into paid work time, but family care and personal care could be affected. In particular, home computer users could be stealing from the 56 or more weekly hours we devote to sleep. Remember that sleep was one of the major things Americans gave up when they started watching TV. History may be repeating itself with the Internet and personal computers.
TAKING IT FURTHER The 1997 SPPA was conducted by random-digital-dial telephone with a national probability sample of 12,376 respondents aged 18 and older, July to October 1997. A household cooperation rate of 65 percent was achieved, with a final overall response rate of 55 percent. To adjust for the somewhat lower proportions of males, blacks, younger people, and less educated respondents, the data were weighted to 1997 Census Bureau characteristics on these factors, and these weights are applied to the analyses.
The 1995 Times-Mirror Technology Survey was conducted by telephone with a national probability sample of 3,603 respondents aged 18 years or older during May and June 1995. Households were selected using the random-digital-dial method. At least 10 percent of the interviews were conducted on each day of the week, and these day-of-the-week differences are adjusted for in the analyses. To adjust for the somewhat lower proportions of males, blacks, younger people, and less-educated respondents, the data were weighted to 1993 Census Bureau characteristics on these factors, and these weights are applied to the analyses. Many of the questions were exact replicates of a parallel 1994 survey. For more information, contact John Robinson at the University of Maryland; telephone (301) 405-5734; fax (301) 314-6892.