**3. Method**

55 Pre-service ESL teachers who were majoring in TESOL at the University Sains Malaysia (USM) participated in this study. Participants were randomly divided into nine groups (on average six members in within each group). As a prerequisite, they developed Community of Practice [25] the group members and created individual e-portfolios. The course *PET301* (Teaching of English through Literature) and *Google Group* were considered as the face-to-face and online setting, respectively. They were required to post Weekly Journals (RJ) in line with the course contents (domain). Furthermore, participants (members) were required to engage mutually to look over others' RJs, examine with an eye of criticism, and write critical reflection as Discussion Journals (RJ) mentioning individual opinion (practice). Qualitative data to explore challenges was collected from these sources. One questionnaire was used as a means of collecting quantitative data. It consists of three parts that corresponds to the main aspects of this study including the participants' perspectives toward e-portfolios, its contribution in their development and growth. Moreover, participants' perspectives section has two parts namely purpose of creating e-portfolios and general perspectives toward e-portfolios. Development and growth section were constructed with four sections—language, assessment, learning and pedagogy. Participants were asked to provide responses within a five-point range from 5 (Strongly Agree) to 1 (Strongly Disagree). Reliability of this instrument was established through a pilot study. The Cronbach Alpha is shown in **Table 1** (results of Reliability Analysis) which confirms the issues consist high levels of reliability and are well above the cut-off value of 0.70 as suggested by Nunally [26]. Data was collected at the end of the course and then it was reduced by measuring mean and standard deviation, and afterward, presented in tables and graphs. SPSS (version 12) was used for this.

A lecture session was arranged at the beginning of the semester to orient them with the concept and demonstrate the procedures of creating e-portfolio. For the Internet connection, weekly tutorial sessions (one hour each) were arranged at the computer laboratory. The Course Instructor (CI) played the role of *E-moderator* [27]


**Table 1.** *Reliability of the instrument.* when participants were making wrong conceptualization or deviating from the main discussion or even when decision-making was required in debatable issues.

Data collection period was limited within one semester (fourteen weeks). Content of the e-portfolios (Reflection Journals) were considered as the source of data. In addition, nine interviews (each from one group) were accomplished for data triangulation and gain insights of interesting or unexpected findings and understand how and why they came to that particular perspective. Participation was voluntary and group members themselves selected their representative for interviews. Data from the interview also served to explore challenges. By examining relevant contents data was organized, broken into manageable units, synthesized, and reduced under different themes. A coding system was used to single out the participants (e.g., A1-- where A refers to the 'Group A' and 1 is the first participant) and data source (RJ refers to the participants' discussion and In for the interviews). For citation, data code and participants' code were used together (e.g., RJF4- reflection journal of fourth participants of group F, InH5- opinion of fifth participant of group H in interview). For qualitative data analysis, as described by Creswell [28], three general processes were followed: preparing and organizing the data, reducing the data into themes and representing data in discussions. Quantitative data of was reduced by measuring mean and presented in tables and graphs.
