Teaching and Learning Forum 2004 Home Page
Teaching and Learning Forum 2004 [ Proceedings Contents ]
Online learning versus face to face learning: What is the difference?

Richard K. Ladyshewsky
Graduate School of Business
Curtin University of Technology

There has been considerable debate about the use of information technology in higher education and whether the technology delivers good educational outcomes. Unfortunately, there is a paucity of 'controlled research', which examines outcomes in online learning (OL) and face to face learning (FTF). This study compared final grades of post-graduate student performance in nine units offered in both FTF and OL mode over the course of two academic years. The influence of age and gender on student outcomes was also considered in this study. Post-graduate students, on average, did better in the OL mode, although, at the individual unit level there were minimal if any significant differences. Students that completed units in both the OL and F2F mode, on average, had significantly higher grades in the units completed online. Age and gender did not appear to moderate performance in any way except for those students under 33 who did better, on average, in the OL mode. The implications for teaching and learning in virtual mediums are discussed.


The increasing use of information technology in higher education has seen monumental growth in the past 10 years (Allen, 1999). Understandably, the growth of online learning (OL) has come under increasing scrutinisation as more and more of this technology is integrated into educational delivery methods (Alavi, 1994).

To date, much of the research on OL has focussed on the internet as an exciting form of technology that can support learning rather than focussing on whether it actually enhances the learning process itself (Sweeney & Ingram, 2001). It has also been raised whether one should even focus on academic outcomes in OL without considering other social and psychological aspects of the education process (Sweeney & Ingram, 2001). McGrath, (1997-98), for example, noted greater inquisitiveness, expressiveness, risk taking, decreased inhibition, reduction of gender barriers, and increased social connectedness as a result of electronic conferencing.

Much of the criticism of OL stems from the inappropriate use of this technology to support learning. A high quality OL experience, just like a FTF learning experience, requires a pedagogical approach that creates a responsive and creative learning environment. Research by Mioduser, Nachmias, Lahav, and Oren, (2000), for example, and their analysis of 436 randomly chosen educational websites, found that in relation to pedagogy most of these online courses reflected traditional approaches commonly found in text books and CD ROM multimedia. To ensure that there is a pedagogical focus to an online unit a variety of principles, espoused by Chickering and Erhmann, cited in McLoughlin, (2000), should be followed. These principles include: student:teacher contact; active learning techniques; prompt feedback; communication of high expectations; time on task; respect for diverse learning communities; and reciprocity and collaboration among students. Arbaugh (2000) argues that class size is another important factor in the success of OL. Smaller class sizes facilitate integration and the development of community and are more manageable by students and staff.

The importance of pedagogical input into the design and delivery of OL is described by Jasinski, (1998):

Technology does not cause learning. As an instructional medium, online technologies will not in themselves improve or cause changes in learning. What improves learning is well-designed instruction. Online learning environments have many capabilities and the potential to widen options and opportunities available to teachers and learners. Technology is coming before pedagogy. ... at this stage of development, the effort put into exploring technologies to 'keep the cutting edge' is at the expense of equal investment in the underpinning of educational design.
While there has been some research comparing FTF and OL directly, these are few in number. The most significant results from good research in this area indicate that outcomes achieved using technology are at least the same as for those in traditional settings (Brennan, McFadden, & Law, 2001). In one study of undergraduate marketing students, Sweeney and Ingram, (2001) evaluated learning preferences for tutorials offered in FTF mode versus online. While web-based approaches were seen as more innovative and enjoyable, the FTF tutorials were seen as more effective learning environments by the students. Web based environments were perceived by the students to have a greater sense of sharing and equality and were more collaborative. However, most students in this study were under 25 and full time. Preference for FTF stemmed from the opportunity for greater interaction with the tutor and the possibility of getting direct information on 'right and wrong' answers.

In another study, Alavi, (1994) compared the performance of a group of Master of Business Administration students in an information systems unit who used computer based group decision support software as part of a collaborative learning exercise. A similar group did the same exercise without the use of this support software. Interestingly, the students' affective reactions to the computer-mediated process were more positive than the control group who did not have access to this technology. Final course grades were also significantly higher for those students that were exposed to the computer mediated discussion environment.

Gender differences in OL have also been investigated in business education by Arbaugh, (2000). Arbaugh's research found that there were no gender based differences in achievement at the post graduate level in a particular business unit. While their sample size was small, they found that women displayed more collective and individual participation patterns than men even though chi-square analysis revealed only moderate significant differences between the genders. They also found that the lack of FTF interaction in an asynchronous OL environment where collaboration is encouraged, does not necessarily lead to a reduction in academic achievement. Richardson and Turner, (2000) found similar gender based findings in their study of OL, which involved a larger sample population. They found, however, that women were more negative towards OL. In this study women self rated as being less computer literate and found the OL environment to be lacking in collaboration. Hence, female preferences for more interactive and collaborative learning outcomes, and computer literacy were not replicated in this study. However, the results do support the findings of Arbaugh, (2000) who noted the importance of collaboration in OL, in particular, for female students.

Other studies have been directly critical of the OL experience. Hara and Kling, (1999) for example, studied the learning experience of six inexperienced computer users studying in an online environment. They found that the learners had three common frustrations: lack of prompt feedback; ambiguous instructions on the web; and technical problems. These same frustrations and problems with OL have also been noted by others in the literature (Arbaugh, 2000).

Meisel and Marx, (1999) in a review of the literature found that OL was less rich in that there was less eye contact, increased emotional detachment and a greater predilection to hold onto strongly held beliefs. In contrast they found that OL was a better facilitator of idea generation in group decision making. Jiang and Ting, (1998) conducted a study of factors that influenced students' OL experience. Their results indicated that the percentage of grade allocation to online discussion and the instructor's specification of requirements for student contributions in the discussion room were significantly and positively correlated to students' perceived amount of learning. Although the level of instructor participation was not significantly correlated with the students' perceived amount of learning, Jiang and Ting found it had a significant correlation with level of students' participation.

Bacani and Rohlfs (2000) noted that OL requires considerable self-discipline. Students who elect to take an OL unit may possess a higher degree of autonomy and motivation from their counterparts who select FTF courses. Dunlop and Scott, (2001) in a study of distance education students found that these students preferred OL significantly over the traditional paper based method. This finding has been paralleled by Richardson and Turner, (2000) who found when they evaluated task orientation and perceptions of OL, students who were inclined to engage with and enjoy independent learning activities had more positive perceptions of OL.

Self-selection or learning styles, therefore, are another factor that may influence preferences for FTF versus OL learning. Halsne and Gatta, (2002) conducted a study of OL and FTF learning at the community college level. There were over 1600 respondents across these two learning examples. In addition to other demographic variables, the researchers evaluated participants' learning styles using Barsch's Learning Style Inventory. The researchers found that participants in OL were predominately visual learners and spent on average an extra hour per week on classwork in comparison to their FTF counterparts. FTF learners were predominately auditory or kinaesthetic learners.

This issue of self selection and learning style was also seen in a study by Felix, (2001). In this study Felix found that students appeared to prefer the mode of study they were 'used to' rather than considering other possible solutions. In their study, distance learners expressed a preference for OL and for using materials on their own. In contrast, FTF learners were not expecting to learn electronically and expressed a desire to maintain the traditional form of teaching. This preference for learning has also been corroborated in other research (Ladyshewsky & Nowak, 2000).

In this research, further evidence is entered into the research pool on OL. This study was a large scale quantitative evaluation of FTF and OL which took gender and age into consideration.


Post-graduate student grades in nine units were captured for two years at the Graduate School of Business, Curtin University of Technology. Data was collected for both the FTF and OL versions of these nine units. All nine units are considered 'fully online', as defined by Finder and Raleigh (1998), involving no face to face interaction with all course content, assignments and communication taking place online.

All nine OL units followed the constructivist learning principles espoused by Chickering and Erhmann cited in McLoughlin, (2000). Each unit is composed of 12 modules or weeks of instruction and contained lecture notes, figures/diagrams, readings and internet links, practical activities and self-assessments, quizzes and on line discussion activities. All modules are moderated by an on line tutor and delivered using IBM LearningSpace software.

The Graduate School of Business operates on a trimester system with the 9 units being offered in different trimesters. Class sizes ranged from 10-40 students in the OL mode and from 15-40 students in FTF mode. Computer literacy of these students was considered to be high given earlier research on the student population (Ladyshewsky & Nowak, 2000).

The specific data that was collected on all students included: gender, age, course enrolment, unit enrolment, and final grade (percentage). Student data was obtained through the receipt of customised reports from the University Statistics Office.

Data analysis

Effect size indicators were used as measures of practical significance (Glass, McGraw, & Smith, 1981; Nelson, 1981). The effect-size method evaluates the difference between the means of pairs of treatment conditions and is best divided by the composite group standard deviation thus yielding a standardised mean difference (Albanese & Mitchell, 1993). Cohen (1969) classifies effect sizes as small (.20), medium (.50) and large (> or = .80). The t test is an inferential test that measures whether random sampling alone is the reason for group differences (Nelson, 1981). These tests are carried out where practical significance is noted and the investigator is interested in ruling out the possibility that chance alone is the reason for the experimental effects. The analysis of variance test was applied to those students that experienced both OL and FTF as part of their course of study. The analysis of variance examines the significance of the differences among two or more groups (Vockell & Asher, 1995).


Table 1 provides information on enrolments across all nine units by OL and FTF categories. The majority of students studied in the face to face environment (77.5 per cent).

Table 1: Enrolments: total, online and face to face

All140131622.5 108577.5
52822182712.4 19187.6
56971822915.9 15384.1
56982894816.6 24183.4
56992365021.2 18678.8
5906902426.7 6673.3
5934361644.4 2055.6
64311636841.7 9558.3
6439802632.5 5467.5
66491072826.2 7973.8

Table 2 provides information on enrolment by gender across all units by OL and FTF categories. Males represented, on average, just over two thirds of the total enrolment (68 per cent). This same pattern was reflected, to a slightly greater or lesser degree in each of the units.

Table 2: Enrolments by gender: total, online and face to face

M%F %M%F%M %F%
All95368448 3220665.211034.874768.833831.2

Table 3 provides information on enrolment by age across all units by OL and FTF categories. The breakdown is for students under 33 years of age and those 33 and older as this represented the midpoint of the age distribution. For the most part age was split relatively equally across the units with a few exceptions.

Table 3: Enrolments by age: total, online and face to face

All71951.3682 49.715749.715950.3 56251.852348.2

Table 4 provides details as to whether or not there are any differences in student performance across the two modes of learning. When taking the average of all student grades across all nine units, students in the OL mode did significantly better. The effect size indicator of 0.1068 suggests small practical significance. At the individual unit level, most units demonstrated no significant difference in either learning mode. There were two exceptions. In unit 5697 students did significantly better in the OL mode with an effect size indicator of 0.505 suggesting moderate practical significance. In unit 6431 students did significantly poorer in the OL mode with an effective size indicator -0.3850, which is of moderate practical significance.

Table 4: Differences in online vs. face to face

Unit codeESIt-valueDeg. of freedomP-valueSig.
All units0.10682.414213990.0161**
56970.5054.1744 1800.0001***
56980.01050.0939 2870.9252NS
59060.27911.5866 880.1162NS
66490.24541.6369 1050.1046NS
** significant at 5% level, *** significant at 1% level and NS means not significant

There did not appear to be any noteworthy differences in academic achievement across the OL and FTF modes by gender. Similarly, there were no significant differences in academic achievement across OL and FTF mediums for individuals 33 years of age and older. There were, however, significant differences across the OL and FTF mediums for students under 33 years of age, which are depicted in Table 5. On average, students under 33 years of age did significantly better in the OL mode when all nine units were taken into consideration. The effect size indicator of 0.1847, in this example, is of small practical significance. This pattern was seen in two of the other units of study, 5697 and 5906, which had moderate and strong measures of practical significance as suggested by their effect size indicators of 0.4797 and 0.6047 respectively. Students under 33 years of age did significantly poorer in unit 6431 with a moderate effect size indicator of -0.3583.

For students that experienced both OL and FTF mediums as part of their study (n=138), an analysis of variance was carried out to determine if their performance differed across these two mediums. There were significant differences (F=7.7, df=1, df=334, P-value=0.006) in the students' performance between the two modes of study at the one percent level of significance with the average score for the OL medium being significantly higher.

Table 5: Comparisons of scores for OL and FTF for students under 33 years of age

Unit codeESIt-valueDeg. of freedomP-valueSig.
All units0.18473.1028 7170.0021***
52820.07450.496 1110.6233NS
56970.47972.5228 1000.0132**
5698-0.0825-0.5361 1390.5927NS
5699-0.0394-0.2266 1200.8212NS
59060.60472.8907 420.0066***
5934-0.2906-0.8629 160.4009NS
6431-0.3583-2.3232 820.0226**
64390.07750.3643 430.7175NS
66490.47671.9657 480.0551NS
** significant at 5% level, *** significant at 1% level and NS means not significant


By focussing on the pedagogy behind unit design and delivery (Mioduser et al., 2000), the academic program in this research study was able to deliver OL units that produced a high quality outcome. A responsive and creative learning environment was designed in keeping with the principles espoused by Chickering and Erhmann, cited in McLoughlin, (2000) and was achieved by ensuring high quality regular contact between students and instructors. Active learning strategies such as self assessments, practical activities, project based assignments and discussion rooms assisted in the engagement of students. It was assuring to see that overall, students in the OL units did significantly better on average than their counterparts in the FTF environment. This concurs with other research comparing OL to FTF learning (Alavi, 1994). At the individual unit level, however, differences in academic performance across the two modes of learning were insignificant and is consistent with research in this area (Brennan et al., 2001). Keeping OL discussion rooms manageable with respect to size was perhaps another factor that ensured a positive educational outcome. When student enrolments within an OL unit approached 20 or more, discussion rooms were split into smaller groups of approximately 10-15 students and managed separately by the instructor. Arbaugh (2000) noted that class size was an important factor in the success of OL as smaller, more manageable groups, appear to facilitate integration and the development of a learning community.

Gender did not appear to have any influence on academic performance across the two modes of learning, which is consistent with earlier research by Arbaugh, (2000). Age was a factor in performance for students under 33 years of age. These students did significantly better, on average, in the OL mode even though the difference is of small practical significance. Arbaugh (2000) suggested that computer literacy was an important determinant of a student's OL experience. These students, because of their youth and exposure to information technology throughout their education, may be more adept learning online.

Perhaps the most interesting outcome from this research was the academic achievement of students who had taken units in both the OL and FTF mode. These students' grades were significantly better in the OL mode. It is not possible from this research to determine if there was a self-selection factor in operation. Self selection issues identified by Halsne and Gatta (2002) and Felix (2001) may explain some of the differences in this study, particularly if distance students and those with a preference for individualistic and visual forms of learning opted for the OL mode (Richardson & Turner, 2000). As noted by Bacani and Rohlfs (2000) and Dunlop and Scott (2001), OL requires considerable self-discipline, and students who elect to take an OL unit, may possess a higher degree of autonomy and motivation. This may have been a small factor in the favorable outcomes seen for OL.

Some of the assumptions and limitations of this study also need to be qualified. Student grades may also not represent the best indicator to evaluate student performance (Sweeney & Ingram, 2001) as a single measure of learning may not completely capture the content and quality of student outcomes (Arbaugh, 2000). The nine units also are different from one another in terms of their content, and the subtle differences in delivery and assessment. Even though units were delivered using a standard template and structure, these variances cannot be ignored as influences moderating student performance. The amount of individual work is generally greater in the OL mode. Group work is used more considerably in the FTF versions of the same unit. Instructors also differ within and between the units. Tutors of an OL unit may not necessarily be the same person that delivers the FTF unit. These are all factors that have an influence on the study's outcomes.

There was also a large difference in the number of students who undertook the two mediums of learning. Face-to-face participants are the larger sample group in comparison to the OL group and this may have had an influence on the power of the statistical analyses.


This research provides some assurance that student performance is at least as good as, if not slightly better, in the OL mode when compared to the FTF mode. When a high degree of pedagogical thought goes into the design and delivery of OL, and is supported by adequate resources, positive educational outcomes can be achieved by students. Quantitative and qualitative research that looks at traditional as well as non-traditional measures of student performance, in controlled situations, are needed to increase our understanding of learning outcomes in electronic mediums.


Alavi, M. (1994). Computer-mediated collaborative learning: An empirical evaluation. MIS Quarterly, 18(2), 159-174.

Allen, M. (1999). Don't be a troll! Using the Internet for successful higher education. Paper presented at the Higher Education Research and Development Society of Australasia Conference, Sydney.

Arbaugh, J. (2000). An exploratory study of the effects of gender on student learning and class participation in an internet-based MBA course. Management Learning, 31(4), 503-519.

Brennan, R., McFadden, M., and Law, E. (2001). Review of research: All that glitters is not gold: online delivery of education and training. Leabrook, South Australia: Australian National Training Authority, NCVER.

Dunlop, M., and Scott, D. (2001). An examination of the impact of aspects of online education delivery on students. Proceedings AusWeb 2001, Southern Cross University. [viewed 17 Mar 2003] http://ausweb.scu.edu.au/aw01/papers/refereed/dunlop/paper.html

Felix, U. (2001). A multivariate analysis of students' experience of web based learning. Australian Journal of Educational Technology, 17(1), 21-36. http://www.ascilite.org.au/ajet/ajet17/felix.html

Halsne, A., and Gatta, L. (2002). Online versus traditionally-delivered instruction: A descriptive study of learner characteristics in a community college setting. Online Journal of Distance Learning Administration, 5(1). [viewed 17 Apr 2003, verified 13 Jun 2004] http://www.westga.edu/~distance/ojdla/spring51/halsne51/

Hara, N., and Kling, R. (1999). Students' frustrations with a Web-based distance education course: A taboo topic in the discourse. CSI Working Paper. [viewed 14 Apr 2003] http://www.slis.indiana.edu/CSI/wp99_01.html

Jasinski, M. (1998). Teaching and learning styles that facilitate on line learning: Documentation project, project report. Adelaide: Douglas Mawson Institute of TAFE.

Jiang, M. and Ting, E. (1998). Course design, instruction, and students' online behaviors: A study of instructional variables and students' perceptions of online learning. Paper presented at the Annual Meeting of the American Educational Research Association, San Diego, April.

Ladyshewsky, R. and Nowak, M. (Eds) (2000). Post graduate business education and flexible learning strategies: An analysis of customer perspectives. Perth, Australia: Curtin University Press.

McGrath, C. (1997-98). A new voice on interchange: is it talking or writing? Implications for the teaching of literature. Journal of Educational Technology Systems, 26, 291-297.

McLoughlin, C. (2000). Beyond the halo effect: Investigating the quality of student learning online. Southern Cross University. Proceeedings Moving Online Conference, pp.141-154. [viewed 12 Apr 2003 at http://www.scu.edu.au/schools/sawd/moconf/mocpapers/moc16.pdf. To obtain, see http://www.scu.edu.au/schools/socialsciences/dds/index.php]

Meisel, S. and Marx, B. (1999). Screen to screen versus face to face: experiencing the differences in management education. Journal of Management Education, 23(6), 719-731.

Mioduser, D., Nachmias, R., Lahav, O. and Oren, A. (2000). Web-based learning environments: Current pedagogical and technolgical state. Journal of Research on Computing in Education, 33(1), 55-76.

Richardson, J. & Turner, A. (2000). A large-scale 'local' evaluation of students' learning experiences using virtual learning environments. Educational Technology and Society, 3(4), 108-125. http://ifets.ieee.org/periodical/vol_4_2000/richardson.html

Sweeney, J. & Ingram, D. (2001). A comparison of traditional and web-based tutorials in marketing education: An exploratory study. Journal of Marketing Education, 23(1), 55-62.

Author: Dr. Richard K. Ladyshewsky, Senior Lecturer, Graduate School of Business
Curtin University of Technology, 78 Murray Street, Perth Western Australia 6000
Email: ladysher@gsb.curtin.edu.au

Please cite as: Ladyshewsky, R. K. (2004). Online learning versus face to face learning: What is the difference? In Seeking Educational Excellence. Proceedings of the 13th Annual Teaching Learning Forum, 9-10 February 2004. Perth: Murdoch University. http://lsn.curtin.edu.au/tlf/tlf2004/ladyshewsky.html

[ Proceedings Contents ] [ Teaching and Learning Forum 2004 ]
This URL: http://lsn.curtin.edu.au/tlf/tlf2004/ladyshewsky.html
Created 13 June 2004. Last revision: 13 June 2004.