The impact of computer literacy on student academic performance in the introductory management information systems course.

Ye, L. Richard

COMPUTER LITERACY AND PERFORMANCE

"Computer literacy" is a commonly used term in the business world, but it is not precisely defined. Computer literacy, in general, is being knowledgeable about the computer and its applications (Rochester & Rochester, 1991). Such knowledge appears to have two dimensions: conceptual, and operational (Winter, Chudoba, & Gutek,1997). The conceptual dimension includes an understanding of the inner workings of a computer or general computer terminology. Without such knowledge a user would find it difficult to figure out any system problems, or to learn to adapt quickly to new systems or software. The operational dimension refers to the necessary skills a user acquires, through training and practice, in order to operate specific systems to complete specific tasks.

While prior research did not evaluate the performance impact of computer literacy empirically, there is evidence that such a performance impact is likely to be task-dependent (Goodhue & Thompson, 1995; Lonstreet & Sorant, 1985; Rhodes, 1985; Thompson, Higgins, & Howell, 1994). For example, if we considered a student to be highly computer literate because s/he demonstrated a high level of proficiency in using a word processor or a spreadsheet program, we would also expect the student to perform well on tasks involving the use of a word processor or a spreadsheet program. We could not predict, however, how the student would perform on tasks involving the use of a database program, if s/he had not received training in database software. This leads us to the following hypothesis:

H1: Students' task performance will be positively correlated with their level of computer literacy, if the same type of software is involved in assessing their level of computer literacy and their task performance.

Winter, Chudoba, and Gutek (1997) use the notion of "functional computer literacy" to argue that a user needs both the conceptual and operational knowledge to perform effectively and productively in various white-collar work settings. A truly "computer fluent" user, they contend, does not simply memorize the correct sequence of keystrokes or mouse clicks. Rather, the user must form an internal representation of the system's structure and functions. Indeed, there is consistent research evidence that links a user's valid mental models of a system to better task performance (Foss & DeRidder, 1988; Booth, 1989; Sein & Bostrom, 1990; Weller, Repman, Lan, & Rooze, 1995). Within the context of computer literacy training, we would therefore expect students to form useful mental models of a computer system based on their conceptual knowledge of the system, and to be able to transfer that knowledge to tasks in an unfamiliar hardware/software environment. This leads us to a second hypothesis:

H2: Students' task performance will be positively correlated with their level of computer literacy, even if the task involves the use of unfamiliar software.

The foregoing arguments suggest that the performance impact of students' computer literacy depends on the nature of the task. More specifically, it depends on whether the task involves the transfer and application of the conceptual and operational knowledge obtained from their computer literacy training and practice. This gives rise to a third hypothesis:

H3: Students' task performance will have no correlation with their level of computer literacy, if the task does not require the use of their conceptual or operational knowledge of the computer hardware/software.

METHOD

A multiple regression analysis was applied to assess the significance of students' level of computer literacy in predicting their task performance. In addition to the primary independent variable, level of computer literacy, the analysis also included two other independent variables: gender, and grade point average.

A number of prior studies have investigated the impact of student gender as a predictor of academic performance, but the results appear inconclusive. Two earlier studies found that female students performed better than males in accounting (Mulchler, Turner, & Williams, 1987; Lipe, 1989), while others found males outperforming females in finance (Borde, Byrd, & Modani, 1996) and Economics (Heath, 1989), but no gender effect in marketing (Borde, 1998).

Extensive research also exists on gender-related computer attitudes and aptitude, with more consistent results. Several studies found that, compared to men, women tend to display lower computer aptitude (Rozell & Gardner, 1999; Smith & Necessary, 1996; Williams, Ogletree, Woodburn, & Raffeld, 1993) and higher levels of computer anxiety (Anderson, 1996; Bozionelos 1996; Igbaria and Chakrabarti 1990). Because the present study focuses on students' performance in an information systems course, we include gender in the research model so its effect can also be explored.

Grade Point Average (GPA) as a predictor of academic performance is widely reported. Numerous studies have found GPA to be significantly correlated with student performance in accounting (Doran, Bouillon, & Smith, 1991; Eskew & Faley, 1988; Jenkins, 1998), marketing (Borde, 1998), and economics (Bellico, 1974; Cohn, 1972). However, because the predictive impact of GPA in an information system course is unknown, we believe it should also be included in the present research model.

This study was conducted among 92 business school students enrolled in four sections of an introductory information systems (IS) course at a public university. All sections were taught by the same instructor, under the same set of conditions. The use of information technology was an integral part of the course requirement. To complete the course successfully, students must complete two hands-on course projects, one hands-on mid-term examination, and a traditional paper-and-pencil final examination.

At the beginning of the course, we measured the students' level of business computer literacy with an existing examination instrument. This provided an individual numeric computer literacy score (CLS), data for the primary independent variable of the study. The instrument had been in use twice a year to determine if a student had met the business school's computer literacy requirement. The exam consisted of three parts: hardware and software concepts, word processing, and spreadsheet modeling. The concepts part was administered on paper in multiple-choices format, while the remaining two parts involved hands-on word-processing and spreadsheet problems to be completed on a computer. Over the course of five years since its first use, the exam had produced a consistent passing rate of 30 to 35 percent. This suggests that the instrument is fairly reliable.

A student information database was used to collect data for the other independent variables: gender, and student GPA prior to taking the IS course. A multiple regression model was run on four performance measures: project 1, project 2, mid-term exam, and final exam. Project 1 involved the development of a database application using a database management system (DBMS), to which the students had no prior exposure. This performance measure was designed to test Hypothesis 2.

Project 2 involved the development of a production plan using spreadsheet modeling. The measure was used to test Hypothesis 1. Students completed the two projects individually, outside the classroom. The hands-on mid-term exam consisted of three parts: IS concept questions answered with a word processor, a database problem, and a spreadsheet problem. This measure was designed to test Hypotheses 1 and 2. The final exam consisted of entirely conceptual IS questions, conducted in paper-and-pencil format. The exam questions focused on traditional IS theories such as transaction processing, decision support, and systems development, which can be considered hardware/software-independent. The measure was used to test Hypothesis 3.

The general regression model was formulated as follows:

[Ps.sub.i] = [$.sub.0] + [$.sub.1]CLS + [$.sub.2]GEN + [$.sub.3]GPA + [sub., i]

Where: [Ps.sub.i] = numeric score on the two projects and the two exams (100 possible on each),

CLS = numeric score on the computer literacy exam (100 possible),

GEN = subject's gender, coded 1 for male and 2 for female,

GPA = subject's cumulative grade point average prior to taking the course (4.0 scale),

0 = intercept,

[$.sub.1-3] = slope coefficient, and

,i = error term.

RESULTS

Descriptive statistics for the various measures of independent and dependent variables are presented in Table 1. The relatively large standard deviation value for CLS suggests that there was a great degree of variation among students' computer literacy levels.

Shown in Table 2 are simple pair-wise correlation coefficients among the independent variables. We found that gender and CLS were negatively correlated at the .05 probability level. This is not surprising. As discussed earlier, prior studies suggest that males tend to demonstrate a higher level of computer aptitude than females. We also found GPA and CLS to be positively correlated; this is consistent with the expectation that high-achieving students make greater efforts in acquiring the necessary knowledge and skills, including computer literacy.

The correlations found in Table 2 do not pose a serious multicollinearity problem. The correlation coefficients were relatively small. We also calculated the variance inflation factors (VIF) for each of the variables. While a VIF considerably larger than 1 would be indicative of serious multicollinearity problems, none of the VIF values calculated for this study was greater than 1.10.

In Table 3, we report the results of the regression analysis. The proposed model appeared to fit well in predicting performance for three of the four performance measures: Project 2, the Midterm Exam, and the Final Exam. Reported coefficients of determination (R2) were 0.21, 0.36, and 0.27, while F values were 9.33, 18.30, and 12.18, respectively, all at a significant 0.01 probability level.

Students' level of computer literacy, represented by CLS, proved to be significant in predicting student performance on Project 2 (t = 2.79, p < 0.01) and the Mid-term Exam (t = 2.38, p < 0.05). These results lend support to the hypothesis (H1) that students' level of computer literacy may influence their task performance, if the task involves the use of software familiar to them. We did not find CLS to be predictive of student performance on Project 1. Therefore there was no evidence to support the hypothesis (H2) that students might be able to transfer their computer literacy knowledge to tasks involving the use of an unfamiliar type of software. This also seems to explain the difference between the t statistic for Project 2 and the t statistic for the Mid-term Exam, since one part of the exam involved the use of database software. Finally, we found no CLS impact on student performance on the Final Exam. This result provides support for H3, that students' level of computer literacy will have no effect on their task performance, if the task requires neither conceptual nor operational knowledge of a computer system.

Of the other two independent variables, we found GPA to be a significant predictor of performance across all four dependent measures. This is consistent with prior research asserting the validity of using students' past GPA as a strong predictor of future academic performance. To high academic achievers, an IS course, despite its heavy technological content, does not appear more difficult than courses in other subject areas.

Overall, gender was not a statistically significant performance predictor for an IS course. Gender's significant effect on Project 2 performance is worth further exploration, however. Because Project 2 involved the use of a spreadsheet program, the result seems to raise an interesting question: are females better at using a spreadsheet program than males? More research is needed to confirm and explain this observation.

DISCUSSION AND CONCLUSION

In this study we examined the performance impact of students' computer literacy in an information systems course. Results of multiple regression analysis found the predictive power of computer literacy to be task-dependent. As expected, if a task requires substantial uses of computers and a specific type of software, students with a higher level of computer literacy, as measured by their proficiency in using that specific combination of hardware/software, achieved significantly better performance. Conversely, if the task requires the use of unfamiliar software, or if the task requires neither the use of computers nor conceptual knowledge of computer hardware/software, students' level of computer literacy had no significant impact on their performance. Instead, students' GPA appeared to have far more predictive power for such tasks.

The study's failure to find support for H2 requires further explanation. Project 1, designed to test H2, required students to develop simple database applications. While the database software was unfamiliar to the students, we had expected them to use their mental models of a computer system to transfer existing knowledge to a new, novel task. What affected their performance, however, did not appear to be their operational knowledge of the software, but rather their understanding of the database concepts. A closer look at their submitted work revealed that many students had difficulty grasping fundamental concepts such as relationships and Boolean logic in complex database queries. While the graphical user interfaces in today's software environment provide great operational consistency across different applications, what seems to dictate task performance, as implied by this study, is an understanding of the task itself. Computers are only tools. A poor understanding of the task will lead to an ineffective use of tools.

The research reported here is limited by several factors. The participants were all from courses taught by one instructor at one university. In the absence of a standard test instrument, the computer literacy examination used in the study was a choice of convenience. While extraneous factors such as instructor styles and task designs were controlled, the results of the study may not be generalizable to different institutions. Further research is needed to overcome these limitations. First, a standard evaluation instrument for computer literacy must be developed and validated. Second, the current study needs to be replicated under different settings.

The findings from this study have important implications on teaching and learning. Despite the universal requirements for computer literacy among academic institutions, the performance impact of such requirements is far from clear-cut. The relationship between a technology-induced increase in students' personal productivity and their academic performance appears, at best, indirect. The predictive power of computer literacy on performance does not depend on whether and how much information technology is used in the completion of a course. Rather, it depends on how much that technology is integrated into the evaluation of student performance. Unless we can demonstrate a clear relationship between computer literacy and performance, many students will continue to perceive the requirement as an unnecessary roadblock to their progress in their chosen academic programs.

Future research also needs to explore the linkage between computer literacy and personal productivity, and the linkage between productivity and academic performance. Establishing the predictive validity of computer literacy requirement is paramount. As information technology is further integrated into the educational process, we need evidence of such predictive validity to help influence students' attitude and behavior toward fulfilling the requirement. Meanwhile, this would also heighten the need for adequate student training in the uses of information technology, if we expect them to be successful.

REFERENCES

Anderson, A.A. (1996). Predictors of computer anxiety and performance in information Systems. Computers in Human Behavior, (12)1, 61-77,

Bellico, R. (1974). Student attitudes and undergraduate achievement for economics majors. Journal of Economic Education, (5)2, 67-68.

Booth, P. (1989). An Introduction to Human-Computer Interaction, Hillsdale, NJ: Lawrence Erlbaum Associates.

Borde, S.F. (1998). Predictors of student academic performance in the introductory marketing course. Journal of Education for Business, 73(5), 302-306.

Borde, S.F., Byrd, A.K., & Modani, N.K. (1996). Determinants of student performance in introductory corporate finance courses, Presented at the Southern Finance Association (SFA) Annual Meeting, Key West, Florida.

Bozionelos, N. (1996). Psychology of computer use: XXXIX. Prevalence of computer anxiety in British managers and professionals. Psychological Reports (78), 995-1002.

Cohn, E. (1972). Students' characteristics and performance in economic statistics. Journal of Economic Education, (3)1, 106-111.

Doran, B. M., Bouillon, M. L., & Smith, C. G. (1991). Determinants of student performance in Accounting Principles I and II. Issues in Accounting Education, (6)1, 74-84.

Eskew, R. K., & Faley, R. H. (1988). Some determinants of student performance in the first college-level financial accounting course. The Accounting Review, (63)1, 137-147.

Foss, D.J., & DeRidder, M. (1988). Technology transfer: On learning a new computer-based system. In J.M. Carroll (Eds.), Interfacing Thought (pp. 159-183), Cambridge, MA: The MIT Press.

Goodhue, D.L., & Thompson, R.L. (1995). Task-technology fit and individual performance. MIS Quarterly, (19)2, 213-236.

Heath, J.A. (1989). Factors affecting student learning - An econometric model of the role of gender in economic education, Journal of Economic Education, (20)2, 226-230.

Igbaria, M., & Chakrabarti, A. (1990). Computer anxiety and attitudes toward microcomputer use. Behavior and Information Technology, (9)3, 220-241.

Jaderstrom, S. (1995). Technology's impact on computer and business curricula. In N.J. Groneman & K.C. Kaser (Eds.), Technology in the classroom (pp. 1-9). National Business Education Yearbook, No. 33. Reston, VA: National Business Education Association.

Jenkins, E.K. (1998). The significant role of critical thinking in predicting auditing students' performance. Journal of Education for Business, (73)5, 274-279.

Jones, M. C., & Pearson, R.A. (1996). Developing an instrument to measure computer literacy.

Journal of Research on Computing in Education, (29)1, 17-28.

Lipe, M.G. (1989). Further evidence on the performance of female versus male accounting students. Issues in Accounting Education, (4)1, 144-152.

Lonstreet, W.S., & Sorant, P.E. (1985). Computer literacy - Definition? Education Horizons, (63)3, 117-120.

Mutchler, J.F., Turner, J.H., & Williams, D.D. (1987). The performance of female versus male accounting students. Issues in Accounting Education, (2)1, 103-111.

Rhodes, L.A. (1985). Moving beyond computer literacy. Educational Leadership, (42)5, 88-89.

Rochester, J. & Rochester, J. (1991). Computer for People: Concepts and Applications. Homewood, IL: Irwin.

Rozell, E.J., & Gardner, W.L. (1999). Computer-related success and failure: A longitudinal field study of the factors influencing computer-related Performance. Computers in Human Behavior, (15)1, 1-10.

Sein, M.K., & Bostrom, R.P. (1989). An experimental investigation of the role and nature of mental models in the learning of desktop systems. In K.M. Kaiser & H.J. Oppelland (Eds.), Desktop Information Technology, North Holland, Elsevier Science Publishers.

Smith, B.N., & Necessary, J.R. (1996). Assessing the computer literacy of undergraduate college students. Journal of Education, (117)2, 188-193.

Tanyel, F., Mitchell, M. A., & McAlum, H. (1999). The Skill set for success of new business school graduates: Do prospective employers and university faculty agree? Journal of Education for Business, (75)1, 33-37.

Thompson, R.L., Higgins, C.A., & Howell, J.M. (1994). Influence of experience on personal computer utilization: Testing a conceptual model. Journal of Management Information Systems, (11)1, 167-187.

Trauth, E.M., Farwell, D.W., & Lee, D. (1993). The IS expectation gap: Industry expectations versus academic preparation. MIS Quarterly, 17(3), 293-307.

Weller, H., Repman, J., Lan, W. & Rooze, G. (1995). Improving the effectiveness of learning through hypermedia-based instruction: the Importance of learner characteristics. Computers in Human Behavior, 11(3-4), 451-465.

Williams, S.W., Ogletree, S.M., Woodburn, W., & Raffeld, P. (1993). Gender Roles, Computer Attitudes, and Dyadic Computer Interaction Performance in College Students. Sex Roles, (29)7-8, 515-525.

Winter, S.J., Chudoba, K.M., & Gutek, B.A. (1997). Misplaced resources? Factors associated with computer literacy among end-users. Information & Management, (32), 29-42.

Zhao, J. J., Ray, C. M., Dye, L. J., & David, R. (1998). Recommended computer end-user skills for business students by Inc. 500 executives and office systems educators. OSRA Systems Research Journal, 16(1), 1-8.

L. Richard Ye, California State University, Northridge

TABLE 1. Descriptive Statistics

Variable M SD N

Dependent Variables
Project 1 score 67.40 23.68 92
Project 2 score 80.67 15.11 92
Mid-term Exam score 56.43 11.25 92
Final Exam score 55.49 11.31 92
Independent Variables
CLS 62.15 21.16 92
Gender 1.51 0.50 92 *
GPA 2.67 0.48 92

* 45 males and 47 females

TABLE 2. Simple Correlation Coefficients

Variable CLS Gender GPA

CLS 1.00 -.24 * .24 *
Gender 1.00 .10
GPA 1.00

* p < 0.05.

TABLE 3
Results for Regression Models for Various Performance Scores:
Standardized Beta Coefficients (t Statistics) for Independent
Variables

Variable Project 1 Project 2

CLS -0.12 0.28
 (-1.05) (2.79) **
Gender 0.05 0.27
 (0.42) (2.66) **
GPA 0.32 0.20
 (2.96) ** (2.07) *
Intercept 30.92 34.83
 (1.88} (3.71) **
Model statistics
Adj. [R.sup.2] 0.07 0.21
F value 3.33 * 9.33 **

Variable Mid-term Exam Final Exam

CLS 0.24 0.12
 (2.38) * (-1.23)
Gender -0.09 -0.04
 (-1.04) (-0.47)
GPA 0.53 0.50
 (6.03) ** (5.36) **
Intercept 19.50 21.52
 (3.13) ** (3.20) **
Model statistics
Adj. [R.sup.2] 0.36 0.27
F value 18.30 ** 12.18 **

* p < 0.05.

** p < 0.01.