Background: Web-based technologies are increasingly being used to create modes of online learning for nurses but their effect has not been assessed in nurse education.

Objectives: Assess whether participation in face-to-face discussion seminars or online asynchronous discussion groups had different effects on educational attainment in a web-based course.

Design: Non-randomised or quasi-experimental design with two groups-students choosing to have face-to-face discussion seminars and students choosing to have online discussions.

Setting: The Core Methods module, of a postgraduate research methods course.

Participants: All 114 students participating in the first 2yr during which the course teaching material was delivered online.

Outcome: Assignment mark for Core Methods course module.

Methods: Background details of the students, their choices of modules and assignment marks were collected as part of the routine course administration. Students' online activities were identified using the student tracking facility within WebCT. Regression models were fitted to explore the association between available explanatory variables and assignment mark.

Results: Students choosing online discussions had a higher Core Methods assignment mark (mean 60.8/100) than students choosing face-to-face discussions (54.4); the difference was statistically significant (t = 3.13, df = 102, p = 0.002), although this ignores confounding variables. Among online discussion students, assignment mark was significantly correlated with the numbers of discussion messages read (Kendall's tau(b) = 0.22, p = 0.050) and posted (Kendall's tau(b) = 0.27,p = 0.017); among face-to-face discussion students, it was significantly associated with the number of non-discussion hits in WebCT (Kendall's tau(b) = 0. 19, p = 0.036). In regression analysis, choice of discussion method, whether an M.Phil./Ph.D. student, number of non-discussion hits in WebCT, number of online discussion messages read and number posted were associated with assignment mark at the 5% level of significance when taken singly; in combination, only whether an M.Phil./Ph.D. student (p = 0.024) and number of non-discussion hits (p = 0.045) retained significance.

Conclusions: This study demonstrates that a research methods course can be delivered to postgraduate healthcare students at least as successfully by an entirely online method in which students participate in online discussion as by a blended method in which students accessing web-based teaching material attend face-to-face seminar discussions. Increased online activity was associated with higher assignment marks. The study highlights new opportunities for educational research that arise from the use of virtual learning environments that routinely record the activities of learners and tutors.