This chapter describes research methods employed in the current study, consisting of the following sections: research design, participants, research instruments, as well as data collection and data analysis procedures.
Research Design
The present study intends to investigate the differences in EFL learners’ WTC in English among four functions of social media and the effect of the selected
personality variable (i.e., risk-taking propensity) on WTC via these functions. It also explores the factors that learners perceive to influence their choices of media to communicate in English.
A mixed method design that involves a questionnaire survey and interviews was used to answer the following research questions:
(1) What is the effect of the social media’s functions on an EFL individuals’
willingness to communicate in English? More specifically, we address the following two questions: (a) Is there a significant difference in an EFL
individuals’ willingness to communicate in English through different functions of the social media in the context of Taiwan? (b) Is there a significant interaction effect of immediacy and communication mode of social media on willingness to communicate?
(2) Does the personality factor of risk-taking propensity have a significant effect on EFL individuals’ willingness to communicate in English through different functions of the social media in the context of Taiwan?
(3) What factors do EFL learners perceive to influence their choice of different social media for communication in English in the context of Taiwan?
Participants
The participants were 197 English-major freshman students from five universities in northern Taiwan. In this way, the researcher hopes to eliminate the influence the variables of age and English proficiency level difference. Most of them (92.8%) have not lived in English-speaking countries before 15 years old for more than three months. Basically, they have studied English for at least six years—three years in junior high and three years in senior high before entering to college, that is to say, they had generally experienced the traditional English-learning process under Taiwan’s formal educational system.
Besides, they were voluntarily involved in the first-stage research, questionnaire survey. All participants approved their understanding toward the two social media in question, Facebook and Skype.
Instruments
In the following sections, the construction and revision of the questionnaires to be utilized in the study are introduced. The questionnaires were designed to measure risk-taking propensity, WTC through different functions of social media, and media choice. For the sake of efficiency and ease of comprehension, all questionnaire items were written in Chinese (see Appendix 1).
It is noted that due to the lack of WTC scale specifically designed for social media use, a WTC-on-social-media scale was constructed by the researcher based on an existent WTC scale. Besides, the researcher revised Ely’s (1986) available scale to measure individuals’ risk-taking propensity. Two items on users’ media choice with respect to communication modes and immediacy were written in the multiple-choice format. Details are provided in the following sections. Before the questionnaires were distributed to the participants, three TESOL MA students were invited to give
comments on the scale and to help the researcher to avoid confusing and ambiguous descriptions.
Social Media Use Questionnaire
At the very beginning, freshman students who were not familiar with the two social media under investigation—Facebook and Skype—were filtered out in order to enhance the validity of the current study. After indicating their understanding of these social media, they were required to reveal how frequent they use these tools in forms of two multiple-choice items.
Background Questionnaire
The background-information questionnaire includes six questions. They are concerned with students’ gender, age, school and major, years of English learning, previous experience of living in English-speaking countries, and their self-rated English proficiency in four skills (i.e., listening, speaking, reading, and writing).
Risk-taking Propensity Questionnaire
With an aim to investigate EFL undergraduate students’ risk-taking propensity, the researcher adopted Ely’s (1986) Language Risk Taking scale items. Ely (1986) conceptualized the construct as:
a lack of hesitancy about using a newly encountered linguistic element; a willingness to use linguistic elements perceived to be complex or difficult; a tolerance of possible incorrectness or inexactitude in using the language; and an inclination to rehearse a new element silently before attempting to use it aloud.
(p.8)
Accordingly, a six-item scale was developed, which include the following items: (1) I
like to wait until I know exactly how to use a Spanish word before using it; (2) I don’t like trying out a difficult sentence in class; (3) At this point, I don’t like trying to express complicated ideas in Spanish in class; (4) I prefer to say what I want in Spanish without worrying about the small details of grammar; (5) In class, I prefer to say a sentence to myself before I speak it; and (6) I prefer to follow basic sentence models rather than risk misusing the language.
To measure students’ risk-taking propensity concerned with English learning in general rather than in the classroom settings, the following modifications were made : (1) replacing “Spanish” in items 1, 3, and 4 with “English”; and (2) omitting “in class”
in items 2, 3, and 5 (see Appendix 2). Similar to the WTC-on-social-media
questionnaire, a four-point response format was adopted. After reversing the score in item 4, all of the item scores were summed up, with a higher total score indicating lower risk-taking propensity.
Willingness to Communicate (WTC) on Social Media Questionnaire
Since no WTC scales on social media was available, the author developed a tailored-made WTC-on-social-media questionnaire based on McCroskey and Richmond’s (1987) WTC scale, which was proved to be highly reliable and valid (McCroskey, 1992). The original scale items address two dimensions: (1)
communication contexts (i.e., in public, in meetings, in small groups, and in pairs) and (2) receivers (including strangers, acquaintances, and friends) and there are a total of 12 items (4 contexts × 3 receivers). The first type of receiver (i.e., strangers) was deleted in the present study since the opportunity to communicate with strangers through social media is arguably low. Following the same logic, the first and the second kind of communication context (i.e., in public and in meetings) were omitted in the revised scale in that communication through social media tends to be privately
conducted and that college students arguably have few chances to have a formal large-scale meeting via social media. With communication avenues afforded by social media as the third dimension of WTC (i.e., Facebook Messenger, Facebook Messages, Skype video calls, and Skype video messages), a 16-item four-point Likert-type WTC scale was developed to elicit learners’ responses (i.e., 2 contexts × 2 receivers × 4 functions of social media). Table 2 below shows the distribution of the items.
Table 2
Different Categories of the WTC-on-Social-Media Questionnaire Items Items
1 – 4 synchronous written Facebook Messenger
5 – 8 asynchronous written Facebook Messages
9 – 12 synchronous oral/aural Skype video calls
13 – 16 asynchronous oral/aural Skype video messages
Items 1 to 4 concern respondents’ WTC on Facebook Messenger as synchronous written-mode of communication; items 5 to 8 concern respondents’ WTC on
Facebook Messages as asynchronous written-mode of communication; items 9 to 12 concern respondents’ WTC on Skype video calls as synchronous oral/aural-mode of communication; items 13 to 16 concern respondents’ WTC on Skype video messages as asynchronous oral/aural-mode of communication (see Appendix 3 for the item statements). Each function’s distinctive features (i.e., their immediacy and
communication mode) were provided in the instruction part to avoid confusion or misunderstanding. The responses range from 1 (strongly unwilling) to 4 (strongly
willing). The higher one’s total score on the scale, the higher one’s WTC via social media.
Media Choice Questionnaire
In addition to the personality effect on individuals’ WTC, the researcher is also interested in other factors’ impact on one’s choice of media with different
communication modes and immediacy. Two items were developed in the form of multiple-choice questions. Besides expressing their preference for each category of social media (i.e., asynchronous versus synchronous; written versus aural/oral), respondents were also requested to indicate reasons for their preferences in the form of multiple-choice questions as well. Also, the meanings of “synchronous” and
“asynchronous” were described in the instruction part to avoid ambiguous interpretation. Options of reasons were developed based on research findings illustrating users’ perceived factors that affect communication in different mode and immediacy (Correa, Hinsley, & De Zúñiga, 2010; Gerbic, 2005; Jepson, 2005;
Joinson, 2004; Short et al., 1976; Watt, Walther, & Nowak, 2002). One open question was provided for respondents to specify reasons not listed in the questionnaire (see Appendix 4 for the items).
Semi-structured Interviews
To elicit more in-depth data, semi-structured interviews will be conducted at the end of the study. Twelve participants who exhibit either a high-level or a low-level of WTC among all functions of the social media were invited to participate in the
interview on a face-to-face basis, via phone call, or via Skype’s phone-call service. As for the criterion to differentiate high-level WTC from low-level one, participants with score higher than the group mean of each function plus 0.5 standard deviation were
labeled as high-level WTC, and participants with score lower than the group mean of each function minus 0.5 standard deviation were labeled as low-level WTC, as revealed in Table 5 in the next chapter.
Specifically, eight high-WTC (average scale score ≧ 3.17, 3.21, 2.90, and 2.96 in the four categories respectively) and four low-WTC (average scale score ≦ 2.68, 2.70, 2.36, and 2.38 in the four categories respectively) students regarding the two social media were interviewed. Focus of the interview was placed on the underlying reasons for their English WTC via different social media. Questions are mainly about:
(1) students’ use of social media; (2) their views about social media of different immediacy and communication modes; and (3) the reasons behind their preferences for social media of different immediacy and communication modes. Each interview lasted for approximately 15 minutes and was recorded.
Data Collection and Data Analysis Procedures The researcher contacted 7 professors at the English departments of 5
universities in the Northern Taiwan by e-mail, seeking their consent for distributing the questionnaire in their English-major freshman classes. The questionnaires were distributed to the students by the researcher and collected as they completed. At the end of the questionnaire, the participants were also invited to provide their phone number or e-mail address for the following possible interviews.
To analyze the collected data, SPSS statistics package for Windows system was used. The results of questionnaire were analyzed through both descriptive and
inferential statistics. To answer the first research question, ANOVA repeated measures were utilized to examine the differences in individual’s WTC among different
functions of social media. To investigate the effect of personality factors (i.e., risk-taking propensity) on WTC through social media, one-way MANOVA
(Between-Within Design) was used. Students were divided into two groups: high risk-takers (average scale score ≦ M – 0.5SD) and low risk-takers (average scale score ≧ M + 0.5SD), based on their scores on the risk-taking questionnaire.
Risk-taking group then serve as the independent factor with WTC via different social medium functions as the dependent factor.
To analyze the qualitative data, the recorded interviews were transcribed and categorized based on the themes emerging from the data.