The Journal of Higher Education ISSN: 0022-1546 (Print) 1538-4640 (Online) Journal homepage: http://www.tandfonline.com/loi/uhej20 College Students' Computer Use Linnda R. Caporael To cite this article: Linnda R. Caporael (1985) College Students' Computer Use, The Journal of Higher Education, 56:2, 172-188, DOI: 10.1080/00221546.1985.11777084 To link to this article: http://dx.doi.org/10.1080/00221546.1985.11777084 Published online: 01 Nov 2016. Submit your article to this journal View related articles Citing articles: 3 View citing articles Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uhej20 Download by: [Florida State University] Date: 28 October 2017, At: 11:05 j-E Linnda R. Caporael Downloaded by [Florida State University] at 11:05 28 October 2017 College Students' Computer Use "A computer for every student!" is the rallying cry for the new alliance between computers and education. Several colleges have announced that students will be required to purchase computers; other colleges have developed incentive plans to encourage purchases. Beyond the predictable financial expenditures, the impact of computers on traditional student life is uncertain. Skeptics expect the present wave of enthusiasm to crash on the shoals of unfulfilled expectations remembered by veterans of computerized instruction . Advocates prophesy "profound impacts" (if the software barrier is surmounted, if there are adequate networks, if there are enough terminals for everyone, if the faculty use them, and so on). And bewildered hoi polloi, with perhaps one or two bouts of computing under their belts, might wonder about the excitement, concluding as did one frustrated student given a computer, "It's a 'do-what-l-say' machine, not a 'do-what-I-mean' machine." Uncertainty about social impacts is inherent in many new technologies, and this dubiety is especially applicable to computers. Current sociocultural concerns, the absence of a shared conception of the computer, and values about the desirability of computers contribute to speculations about their impact . This research was supported by the IBM Corporation. The author gratefully acknowledges the contributions of Valerie E. Orlop, research associate, and of Daniel Zuckerman, Laurie Blake, and Glen Culbertson. Special thanks go to James Moss, director of the Office of Computer Services at Rensselaer; Richard Park, chairman of the Darrin Scholar Steering Committee; and, most of all, to the students who so willingly shared their experiences with us. Linnda R. Caporael is assistant professor oj psychology in the department oj Science and Technology Studies at Rensselaer Polytechnic Institute. Journal oj Higher Education, Vol. 56, No.2 (Marchi April 1985) Copyright Е 1985 by the Ohio State University Press Downloaded by [Florida State University] at 11:05 28 October 2017 Computer Use 173 These predictions, although provocative, tell us more about the present than about the future. The first effect of a new device requiring attention and concentration is the displacement of time once used for other activities. In the case of computers, displacement will depend on the availability of hardware and how its use requires time . Other effects will be a function of how the computer actually is used rather than how it could be used . The replacement uses of a popular innovation are characterized by familiar forms in a new medium. An example is early television advertising, which used a format familiar from radio advertising. Emergent uses are characterized by unpredictable, exploratory forms and uses for a new medium, some of which may become standard usage of the technology. Among many such emergent uses of computers might be computer conferencing. Research on an innovation is complicated by the reciprocal effects between the new device and the social milieu into which it is introduced: existing social factors influence how the computer will be used, and the availability of the computer may change the factors influencing its use. Thus, although we know there will be emergent uses and effects, specifying which of the many experimental variations will become standard is the job of those condemned to study history , who confront the problem of extrapolating from "horseless carriage" the analytic concept of "suburbs" and of foreseeing the demand for multilevel parking lots. One possible way to detect trends is by local experiments undertaken to identify questions, hypotheses, and variables for subsequent investigation in other settings. Meta-analysis of many such experiments - for example, those on the effectiveness of computer-based instruction - may be the only viable strategy for investigating technologically induced change. This research reports one such local experiment on college students' use of computers. Its purpose was to study time allocation associated with owning a personal computer, to investigate how students used computers, and to identify potential educational and social impacts. The study took advantage of a novel scholarship program that awarded microcomputers to academically talented students regardless of their interest in computing. Two comparison groups were recruited, one composed of academically similar students without computers and the other composed of students already owning computers at college entry. Participants in the two-year study reported their activities in a structured diary  for three days each semester and were individually 174 Journal of Higher Education interviewed after at least the first year of college. The interview data are discussed separately. Downloaded by [Florida State University] at 11:05 28 October 2017 Time Allocation Subjects and Procedure The study is designed around two freshman cohorts: Cohort A entered in 1981 and Cohort B entered in 1982. Each cohort is divided into three comparison groups of approximately twenty students per group. One of these is composed of the scholarship program (SP) students, who were selected for their awards on the basis of high-school class standing, Scholastic Aptitude Test (SAT) scores, and clinical judgment. A declared major in computer science or an interest in computing were not factors in their selection. In order to compare the scholarship students with others of similar academic talent, the second group in each cohort was composed of volunteers from a pool of the top forty students having similar SAT scores. These students were the "no personal computer" (NPC) group. And, to be able to compare the scholarship students with other students who also owned microcomputers, the "own personal computer" (OPC) group was composed of students who, regardless of their academic standing, were invited to participate because they owned a personal computer at college entry. SP and NPC students were initially contacted about the study by letter; OPC students were recruited by announcements and word of mouth. The Cohort A-SP students received an Atari 800 microcomputer. The Cohort B-SP students received an IBM personal computer. Over time, as the scholarship students learned more about their computers, they were expected to diverge from the NPC group and become more similar in computer use to the students owning personal computers at college entry. SAT scores were used to test the assumption that SP students and NPC students were academically similar. Because the sample could not satisfy the standard analysis of variance assumptions for normality and homogeneity of variance, the statistical tests are based on random permutation techniques that empirically generate sampling distributions to assess chance probabilities . Randomization testing also permitted the use of a very simple and direct test statistic, the difference between the means (DM). Table 1 shows there was no evidence disconfirming the assumption that SP students and NPC students were similar (Cohort A: Verbal SAT, DM=23.2, n.s.; Math SAT, DM=22.6, n.s. Cohort B: Verbal Computer Use 175 TABLE I Downloaded by [Florida State University] at 11:05 28 October 2017 Means and Difference Tests for High-School SAT Scores Group n Verbal SAT Math SAT Cohort A SP NPC OPC Differences 19 19 17 735.8 712.6 573.7 SP vs. OPC' 755.8 733.2 703.7 SP vs. OPC' Cohort B SP NPC OPC Differences 19 25 23 728.5 704.0 568.3 SP vs. OPC' 771.5 749.6 697.5 SP vs. OPC' NOTES; No significant differences were found between the SP and NPC groups. SP-scholarship program; NPC-no personal computer; OPC-own personal computer. *p<O.Ol. SAT, DM=24.5, n.s.; Math SAT, DM=31.9, n.s.). In both cohorts, for both types of SAT tests, the SP students did score significantly higher than the OPC students (Cohort A: Verbal SAT, DM = 162.1, p<O.OI; Math SAT, DM=52.1, p<O.Ol. Cohort B: Verbal SAT, DM= 160.2, p<O.OI; Math SAT, DM= 74.0, p<O.OI). The statistical difference between students on the math scores was not considered practically significant because the group mean scores are three standard deviations above the normalized mean SAT of 500. The real difference among students lies in their verbal SAT scores. The most frequently reported major in all groups was engineering, followed by computer science, but the curriculum for the first two years is almost identical for all majors. The study's methodological core was students' reports of their allocation of time on ten mutually exclusive activity categories (for example, sleep, study, or computer use). Each semester students completed a three-day sequence, excluding weekends, of structured diaries. (Fall 1981 is an exception because only one day of activities was recorded by Cohort A.) The student indicated his or her activity by drawing a line through fifteen-minute-interva1 markings in the appropriate category. The student reports were staggererd throughout the semester, and effects for particular days were assumed to be random. More specific information was requested for some categories, such as computer use, and was coded into subcategories by the research staff. The coding categories were developed during pilot work. Confidence in the validity of students' reports is moderately high though not complete. It is much higher than it would be for questionnaires. Downloaded by [Florida State University] at 11:05 28 October 2017 176 Journal of Higher Education For clarity in presentation, the reporting categories were lumped a priori into four analytic categories: personal care (sleeping, eating, chores,); social activities (socializing with friends, entertainment, club meetings, sports); academic activities (attending class, studying from books); and computer activities (any computer activity on any computer system). Three qualifications are in order. First, the overall strategy was to explore not to confirm the hypothesis. Second, the research was correlational: differences between groups might be associated with computer ownership but might not necessarily be caused by computer ownership. Third, the extent to which results might be generalizable is a matter of common sense not statistical inference. The setting for this study, Rensselaer Polytechnic Institute (RPI), is considerably more technical than a liberal arts setting. Results General activities. Data for each cohort were analyzed separately. Table 2 shows the means and test results for differences among the groups' activities. Cohort A data are based on students' average times for four semesters; Cohort B data are based on two semesters. As one would expect, data on the NPC students were stable over both cohorts, as were data for all groups on the personal care variable. But the expected similarity between OPC groups in the two cohorts was not obtained. SP students also differed between cohorts, a fact not completely surprising because different microcomputers were awarded in the two years of the scholarship program. The pattern for the allocation of time to computer use was similar for both cohorts, albeit Cohort B-OPC students engaged in less computing than did the correspondTABLE 2 Means and Differences for Activities Academic Computing n Personal Care Social Cohort A SP NPC OPC Differences 21 22 21 10.4 10.8 10.3 5.1 4.6 5.1 5.6 7.2 5.9 SP vs. NPC" 1.5 0.6 2.0 SP vs. NPC" Cohort B SP NPC OPC Di ff erences 19 25 24 10.5 11.0 10.3 3.6 4.7 4.0 SP vs. NPC' 7.4 7.2 6.9 1.7 0.5 1.4 SP vs. NPC" Group NOTES: Results for Cohort A are based on four semesters, Cohort B on two semesters. SP - scholarship program; NPC - no personal computer; OPC - own personal computer. иp<O.05. ииp<O.Ol. Downloaded by [Florida State University] at 11:05 28 October 2017 Computer Use 177 ing Cohort A group. For both cohorts, SP students allocated significantly more time to computing than did NPC students, but SP students showed no significant differences when compared to OPC students. This finding was the case not only for overall computing data but also for the data on the first semester of the freshman year. Within each cohort, there were no significant differences in time allotted to the four activities between the SP and OPC students. SP students in Cohort A, however, did spend less time on academic activities than did the academically similar NPC students; but in Cohort B, SP students spent less time on social activities than did NPC students. One possible interpretation of the overall pattern is that the significance of owning a computer changed between cohorts largely as a function of the diffusion of computing throughout our culture. The data, our knowledge of the research participants, and conversations with older students owning computers suggest that in 1981 (before the IBM personal computer was available), few students owned personal computers. Those who did were highly committed to and experienced with computing, and student owners participated in a tightly knit "computer culture." By the following year, students less committed and experienced with computers owned them, often more from a commitment to competing academically than from a commitment to computing. The new accessibility of computing at a microcomputer lab also diversified members of the "computing culture," reducing potential social obligations. This analysis suggests that entering students owning computers in the two years had different priorities for allocating their time. We took a closer look at the relationship between the report categories and ownership of a microcomputer. Across cohorts, students who owned computers showed a small but significant inverse correlation for time spent studying (excluding computer use) and computing (Spearman's rho=-0.21, n=85, p<0.05), but they showed no correlation between computing and socializing. Students who did not own computers showed an inverse correlation between socializing and computing (Spearman's rho = -0.42, n = 47, p< 0.01), but they showed no correlation between computing and studying. Sociocultural variables are likely to account for the pattern of time use across groups, but the intraindividual data suggest that personal access is associated with a predictable shift in the use of time, a result consistent with Condry and Keith's  prediction that time displacement effects depend on hardware availability. There are many possible reasons for the shift, among them that owning a computer gener- Downloaded by [Florida State University] at 11:05 28 October 2017 178 Journal of Higher Education ates a new fit or adaptation between the student and his or her environment. If this interpretation is correct, we might view students owning computers as living in a new domain, different from the domain of students without computers. Nevertheless, there is still insufficient reason to assume that this new domain is final (in the sense that television, for example, has displaced time from reading) rather than transitory. Computing activity. Students' reports of their computing activities, collapsed across cohorts, were coded into the five categories shown in Figure 1. ("Other" includes miscellaneous activities such as tinkering with hardware or software and sending messages. Also, the breakdown of the computing category was not done for Cohort A in Fall 1981 because one day of data is insufficient for coding into subcategories.) The data are presented in terms of the means for the groups 3.0 ......- - - - - - - - - - - - - - - , DTI SP 2.5 ~NPC III OPC 2.0 >< Q) "C c 1.5 Q) E t- 1.0 0.5 HW G PP WP OTH Activities FIG. 1. Computing Activities of Students. The time index is the average number of hours over a three-day period. Group identifiers are SP - scholarship program; NPC - no personal computer; OPC - own personal computer. Activity identifiers are HW - homework; G - games; PP - personal projects (excludes required classwork); WP - wordprocessing; OTH - other activities. Downloaded by [Florida State University] at 11:05 28 October 2017 Computer Use 179 over a three-day period. Overall, the largest category of use is for homework in computing classes. There were no significant differences in the number of computer classes taken by students until the second semester of the sophomore year when NPC students took fewer classes than did the other groups in Cohort A. After homework, the next most frequent category was games. OPC students account for the most game playing, as well as the greater portion of activity in the other categories. Different computer systems - that is, the mainframe, the personal computer linked via a modem to the mainframe, and the personal computer as a stand-alone system - are clearly associated with different activities. Careful inspection of Figure 2 indicates that, whether or not students own computers, they use the mainframe primarily for homework and they use it at almost identical levels. The extra time allocated to computing by students owning microcomputers occurs on configurations involving the personal computer. The data suggest 3.0....-----------------, ITIIDsp 2.5 ~NPC RoPC 2.0 >< C1l 1:' c C1l 1.5 E ~ 1.0 0.5 HOMEWORK 0.0 HOMEWORK GAMES L-L_.L.-......L.---l........._"'-----'----I........._L...-....L---a........ MTS Linked Systems PC FIG. 2. Primary Uses for Different Computer Configurations. The white area indicates the time index for the primary activity on the system and the hatched area represents all other activities. Group identifiers are SP - scholarship program; NPCno personal computer; OPC - own personal computer. Downloaded by [Florida State University] at 11:05 28 October 2017 180 Journal of Higher Education that microcomputing will not automatically reduce reliance on the mainframe system. On the contrary, microcomputing appears to be associated with higher usage of the mainframe system, with the microcomputer used as a remote terminal improving access to the campus computer. Students do virtually no homework at all on the stand-alone system. The reasons are fairly obvious - most computer courses use languages that are not readily available on most microcomputers, and even when those languages are available, the debugging routines for programming are not. Grades. Grades were not correlated with computing - nor for that matter with studying. They did correlate with SAT scores (Spearman's rho = 0.30, n = 118, p < 0.01), an unsurprising finding since the SAT is designed to predict academic performance. Student Interviews Individual interviews were conducted with 118 participating students (67 freshmen and 51 sophomores) in Spring 1983. The interview schedule consisted of open-ended questions on various topics covering educational and social issues associated with computing. Responses were coded after the interviews were completed. The interviews were preliminary and exploratory; their primary purpose was the identification of variables and issues for future investigation. Because of the small number of students in each of the six groups and of the openended structure of the interview, data were collapsed across groups for analysis. Learning to Use a Computer Students' reports of computing experience before college were consistent with the possibility, suggested by the time allocation data, that differences between cohorts were related to the diffusion of computing through our culture. Only twelve students (10 percent) had no prior computing experience, almost all of them 1981 freshmen. Forty-three percent of the students reported having at least one high-school computing course, and 27 percent reported three or more. The remainder took college credit courses in high school. The most frequently reported activity was "playing around" with the computer, followed by programming. Twenty students (17 percent) felt that their high-school experience had not been helpful in learning to use the mainframe at Rensselaer. The remainder gave assorted reasons for why their high-school experience had been useful: they learned to think logically, operating sys- Downloaded by [Florida State University] at 11:05 28 October 2017 Computer Use 181 tems were similar, or they felt comfortable with computers as a result of their experience. Students reported, in order of preference, a variety of strategies for learning to use the mainframe system. The most frequently mentioned was trial and error (75 percent), followed by taking a class (55 percent), asking people (43 percent), reading a short primer distributed by the Office of Computing Services at RPI (42 percent), and reading the documentation manuals (29 percent). Based on the order in which students mentioned the strategies used (e.g., taking a class was the most frequent strategy in the first position), a profile of the sequence of strategies was constructed. Beginning with the first position, these were taking a class, trial and error, asking people, and reading the documentation manuals. It is widely acknowledged that reference manuals are very difficult to understand, and this fact is the likeliest explanation for their relatively infrequent use. Students were specifically asked if people or books were more important to them in learning to use the computer: 53 percent thought people were more important, and 47 percent thought books were more important. We suspected that students' responses might be related to how sophisticated they were vis-a-vis computing. Knowledge of Assembler (a common but cryptic programming language) was used as a proxy variable for computer sophistication, and student responses were regrouped according to this variable. Among students knowing Assembler (n = 44),61 percent preferred books and 39 percent preferred people. The percentages were exactly reversed for the students ignorant of Assembler (n = 70): 61 percent preferred people and 39 percent preferred books. The reasons given for both preferences were the sameeither people, or books, were faster, easier, and more specific. The likeliest interpretation for these results is that novices learn from other people who are able to infer the question being asked even if it is phrased in nontechnical terms, while sophisticated users, who generally have command of the technical terms, are able to consult a manual. Living with a Microcomputer Seventy-six students had microcomputers in their dorms or apartments. Over half, 57 percent, kept the computer on a desk in their room, and 34 percent reported that this arrangement created a space problem. Among students reporting other locations for their computers (table, shelf, and so on), 15 percent reported a problem. There were six students who stored their computers. Ninety percent of the students claimed to take security precautions to protect their computers, but almost all of them also state they would take the same precautions Downloaded by [Florida State University] at 11:05 28 October 2017 182 Journal of Higher Education to protect the contents of their room if they had no computer. The upshot appears to be that students take no special security precautions for their computers. Interestingly, about half (48 percent) of the students reported the computer helped them to make friends. It attracted other students to their room and helped them meet people with a common interest. Only five students (four of them in the OPC groups) reported that they wouldn't let others use their microcomputer. Among students who would share, 51 percent reported that specific friends used the microcomputer on a regular basis. About a third of the students owning their own computers (SP and OPC) and half of the NPC students used other people's computers. In general, it appears that student-owned computers tend to be a shared resource. The predominant activities among owner-users and friend-users were reported to be games and remote terminal use, but projects, wordprocessing, calculations, and graphics were also mentioned. The computer was used as a break from studying (activity unspecified) for an average of 3.6 hours per week. Personal Computers as Remote Terminals About half the students (58) had modems, devices that allowed them to link with the campus mainframe system. Students report that their microcomputers are used in this way for an average of 8.5 hours per week. This time includes use by the students and by any others whom they allow to use their computers. (Data from the time diaries indicate that research participants use their computers for mainframe access an average of 3.0 hours per week.) But only two students reported that tying up the telephone line created an inconvenience. Thirty-nine students had problems linking to the mainframe. The most frequently mentioned difficulty was telephone line trouble, followed by inadequate software, and busy or no-answer signals on the mainframe telephone. Also mentioned were problems with the students' own equipment and the cost of the phone call. Disadvantages of using the modem were coded separately from the problems students experienced with them. The former is very dependent on the availability of alternative computer systems, and at Rensselaer students have access to "smart" public terminals with a range of desirable capabilities. The disadvantages students report can easily be summarized - a microcomputer with a standard telephone link is not a smart terminal. The most frequently mentioned disadvantages of using microcomputers as remote terminals included the lack of full-screen editing, slow speed when linked, limited screen size, and no program- Computer Use 183 Downloaded by [Florida State University] at 11:05 28 October 2017 mable function keys, scrolling, or page buffer. In general, the problems and disadvantages experienced by students are associated with the limitations of linking over a telephone modem (e.g., lack of full-screen editing) as well as the limitations of the equipment they own (e.g., limited screen size). Student Projects Thirty-two students reported undertaking special projects above and beyond their classroom responsibilities. Seven were unrelated to computing (e.g., historical research, radio construction) or used computers for calculations only. Students who did not own computers did not report computer projects on any system: only computer owners engaged in computer projects. The computer projects covered a wide range of topics and varied considerably in their sophistication. Several students were working on game programs, several others were working on a microprocessor-based controller for a student-run space telescope project. Examples of other projects were a program for helping cerebral palsy victims communicate, software development for doctors, terminal emulators, modifications of existing programs, and developing small demonstration progams (e.g., one that matches people's interests). Project initiation and development is largely a social activity with peers: two-thirds of the students' projects were undertaken with friends. In general, it appears that students who report working alone are engaged in very sophisticated projects (e.g., a two- and three-dimensional imaging program for the IBM personal computer or a Lisp interpreter); whereas the collective projects seem to have a greater range. Only three students, all in the OPC groups, reported approaching professors to initiate a project. The initial data on student projects is promising, especially considering that the students were freshmen and sophomores. It is likely that, as students mature in their academic careers, more of them will initiate projects of wider scope and complexity. Computing and Classwork Only three students reported having absolutely no courses for which they used any computer. Computers were used in noncomputing courses (e.g., science or engineering) by 60 percent of the respondents. Overall, sophomores (Cohort A), on average, had taken about five college courses where computers were used, and the freshmen had taken an average of three. The highest number of reports were for computer Downloaded by [Florida State University] at 11:05 28 October 2017 184 Journal of Higher Education science courses (114) followed by science (usually physics), of which there were 58 reports. Interestingly, there were twice as many reports (23) of computers being used in humanities and social science courses than in engineering courses (11 reports). It is difficult to categorize the classwork activities associated with computing because of the overlap of tasks. With this qualification in mind, student responses were grouped as follows: class projects (86 percent), textprocessing (46 percent), calculating homework problems (20 percent), and plotting (17 percent). More details were requested about the particular activity of writing papers for class. Sixty-seven students reported one or more assigned papers. Of the eighty-four papers written, 31 percent were typewritten, 25 percent were textprocessed on the mainframe, 23 percent were textprocessed on a microcomputer, 19 percent were handwritten, and 2 percent were textprocessed on a microcomputer and printed on the mainframe system. The infrequency of the last category captures a central frustration of the computer age - incompatibility. Ideally, one should be able to prepare the final version of a paper on a personal computer and duplicate the completed version on a mainframe so it can be copied on a high quality printer. But this capability is not available at Rensselaer, and its closest approximation involves mastering two wordprocessing systems. One of the few closed questions of the interview concerned the systems used for different activities, the latter derived from McCredie's Campus Computing Strategies . They represent a panoply of potential uses for a computer in a university environment. Students were shown an activity-by-system matrix and asked to check every applicable cell. The matrix and percentage of students indicating use is shown in Table 3. In general, student activity seems high in all categories, and they engaged in a wider range of activities than those taught in the classroom. Our impression is that knowledge about these activities is spread through a student network. Opinions on the Future of Computing Students are ready and waiting for the inclusion of microcomputers in their college courses: over 95 percent of the students thought microcomputers could be integrated into the classroom. But most students were unable to respond to a probe about how this could be accomplished. Twenty-three percent volunteered that it was important that students either individually own microcomputers or have easy access for self-pacing. Several (12) added that the computer would be a more Computer Use 185 TABLE 3 Downloaded by [Florida State University] at 11:05 28 October 2017 Percentage of Students Indicating System Used for Specific Activity Activity PC PC/MTS Mrs Games (84) Data analysis, calculations, plotting data, statistical package (82) Electronic mail (78) Textprocessing (67) Information retrieval, library database (55) Simulations, modeling (36) Computer controlled experiments (28) Class evaluations, test-taking (15) 66 18 48 3 53 15 31 12 21 57 59 11 4 6 6 7 2 2 20 38 II 17 23 5 45 37 18 3 13 Other - - - - - - ------'''--'----'-------------------- - - ---- NOTES: Numbers in parentheses are the percentage of students indicating they had engaged in the activity; to the central mainframe campus computer. ---- n ~ 118. MTS refers effective educational tool if more people knew how to use it and started learning at a younger age. About a quarter of the students recommended lectures in class or special classes on how to use microcomputers for various tasks. Students were much more specific about what features would be useful to them in a microcomputer. Some of their responses reflect their experience with their own equipment (e.g., having a printer would be useful). Other responses reflect dissatisfaction with circumstances at Rensselaer (e.g., transferring files between systems). Still others reflect the available technology (e.g., user-friendly features). There was little consensus about how the computer would change education. The 118 students listed over 60 different changes, most positive, many neutral, and very few negative. Only three students thought computers would replace the classroom, and five listed dehumanizing effects. By and large, student opinions reflected those often read in the media and were equally contradictory. Some felt there would be less face-to-face interaction; others felt there would be more student-faculty interaction. Some thought there would be more "number-crunching"; others believed in greater possibilities to learn general principles and theories rather than do busy-work. Our impression during the interview was that students, like most people, lacked enough experience with computing on which to base their responses. General Conclusions This study's perspective was that the social and educational impact of computers will depend on the purposes, replacement and emergent, for which computers will be used once they are widely adopted. Thus, Downloaded by [Florida State University] at 11:05 28 October 2017 186 Journal of Higher Education conclusions are tentative at best because the data describe a dynamic process of change and transition rather than a static state where computing is as casual as television watching. By the same token, computing may be diffusing through society as fast or faster than did television, and it is important to generate questions and identify potentially important variables in the early stages and possibly to adopt strategies that may influence the direction of the computer transition. The primary uses of the computer are as replacements. At present, it does seem fair to say that the appropriate analogy for the microcomputer is a typewriter or calculator. These tools contribute to academic productivity but are hardly central features of instructional activity. Currently, the primary use of the microcomputer is as a replacement for a terminal to the mainframe. As a stand-alone system, the primary use of the microcomputer is as a game machine, but this use is quite variable among semesters. Computer games might be classified as an emergent use of computers, and there are [cf. 8] arguments that the structure and requirements of a game may match the characteristics we would desire in educational software (e.g., challenge, the requirement to solve a problem, rapid feedback). Nevertheless, most observers would agree that the potentials of a microcomputer would be lost if the sole use, emergent or otherwise, was for computer games. The many differences in the microcomputer-owning group, compared to the relative stability of the nonowning groups, suggest educational microcomputing is in a transitional stage as students try to integrate the computer into the academic environment. A niche exists for software to support educational activities: the software does not now exist or is inadequate. The inverse correlation between studying and computing observed in groups owning microcomputers suggests that students are already displacing studying with computer use. Moreover, the data show that the extra time spent on computing is for classwork, although it appears that NPC students, who largely have the same curriculum as other students during these early years, are adequately meeting class demands without additional computer time. This finding suggests that a niche for educational computer use is available within the context of students' priorities, but the software for wider use of computing across the curriculum is not available. Aside from curricular software, the student interviews clearly indicate the need for support software. In particular, software that would improve access to the mainframe and increase compatibility between software used on different systems is essential. Another area Downloaded by [Florida State University] at 11:05 28 October 2017 Computer Use 187 of support development is for general purpose educational software, for example, an academic version of a data base management system that is conducive to recording and searching through notes. Learning to use a computer has two important components: the opportunity for trial-and-error learning and the social transfer of information. Students describe their learning strategies as "playing around," "tinkering," and "experimenting." Learning is an active process where students try a command and observe what happens; they seek out the information they need to know. For the novice, socially transmitted information is the most efficient way to find some particular piece of information necessary for this search without requiring that the information be described in technical terms. Students are not able to describe this process in any greater detail, an inability not surprising given the shortcomings of memory. A better understanding of the trial-and-error process and the role of other people in the learning process requires a descriptive research project exclusively focused on the learner. Computing is a social activity. Repeatedly, evidence indicated that the social environment is an enormously important medium for computing activity. Novices reported a high level of dependence on more knowledgeable students for information about the computer. Students reported in their interviews a far wider range of activities (e.g., electronic mail) than those assigned in the classroom, and other students are the most likely source of their knowledge. Two-thirds of the projects students reported were initiated with friends - and even the most superficial projects are a vehicle to learning more about the computer. Individually owned computers are by no means individually operated computers: microcomputers are a shared resource and a focal point for showing off one's knowledge, aiding a friend, cooperating on a group endeavor, and discovering what computers are good for. The findings suggest that, in addition to asking what the impact of computing on students will be, we should also ask what the impact of students on computing will be. Student culture appears to be an important factor in how microcomputers will be used, and this possibility may be especially so when the machines are not part of the educational efforts of the institution. Student computing projects suggest that students may be a vastly underrated and undeveloped resource for inventing emergent uses for the computer. Innovative policy strategies might include those that encourage student applications of the computer (e.g., an equivalent of the school science fair) and increase 188 Journal of Higher Education the opportunities for learning about computers by increasing the "bandwidth" of the social transmission of knowledge (e.g., providing small, publicly available microcomputer clusters). References Downloaded by [Florida State University] at 11:05 28 October 2017 1. Caporael, L. R. "Computers, Prophecy and Experience." Journal oj Social Issues, in press. 2. Caporael, L. R., and W. Thorngate. "Introduction." Journal oj Social Issues, in press. 3. Condry, J., and D. Keith. "Educational and Recreational Uses of Computer Technology: Computer Instruction and Video Games." Youth & Society, 15 (1983), 87-112. 4. Edgington, E. S. Randomization Tests. New York: Marcel Dekker, 1980. 5. Fischoff', B. "For Those Condemned to Study the Past: Reflections on Historical Judgment." In Fa!!ible Judgment in Behavioral Research, New Directions jor Methodology oj Social and Behavioral Science, No.4, edited by R. Shweder , pp. 79-93. San Francisco: Jossey-Bass, 1980. 6. Kulik, J. A., C. Kulik, and P. Cohen. "Effectiveness of Computer Based College Teaching: A Meta-analysis of Findings." Review oj Educational Research, 50 (1980), 525-44. 7. McCredie, J. W. Campus Computing Strategies. Bedford: Digital Press, 1983. 8. Malone, T. W. What Makes Things Fun to Learn? A Study oj Intrinsica!!y Motivating Computer Games. Cognitive and Instructional Science Series, ClS-7, SSL-8-11. Palo Alto: Xerox Palo Alto Research Center, August 1980. 9. Oettinger, A. G. Run, Computer, Run: The Mythology oj Educational Innovation. New York: Collier, 1969. 10. Robinson, J. How Americans Use Time. New York: Praeger, 1977.