Participants in the study were students at a Singapore junior college and four of their English and Literature teachers. In Singapore, students who do well on the Cambridge O-Level exam taken at about age 16 are eligible to attend junior college. These two-year institutions prepare students for the Cambridge A-Level exam which is a key criterion for university admission. Both the O-Level and A-Level are external exams for all major school subjects, including English.

Writing scripts from 181 students (93 female and 88 male) were examined, while 258 students (141 female and 117 male) completed a questionnaire on their views and practices regarding gender-inclusive English. The former group were chosen based on the essay topics they had chosen, as will be further explained in the section on data collection. The latter group of students were gathered by the third author based on having a free slot in their class schedule on the day data were collected. As the questionnaires were done anonymously, it was not possible to know if some of the students whose writing was examined had also completed the questionnaire. Six students, four females and two males, were interviewed. They had been selected by the third author, a teacher at the college, as representing a random sample of students. Four teachers were also interviewed, two females and two males. Three were Singaporean and one was British. They were selected by the third author to represent a cross-section of experienced teachers at the college. Their teaching experience ranged from 7-20 years.

Materials

The 15-item questionnaire (Appendix 1) was an adaptation of one used by Rubin, Greene, and Schneider (1994) to collect data on the views and practices regarding gender-exclusive/inclusive English of students at a U.S. university. The questionnaire used in the present study had two sections. The first was designed to provide data on students' awareness of the issue, practices at their previous and current educational institutions, their views on the fairness of gender-exclusive English, and their level of concern about the issue. The second section of the questionnaire provided demographic data on students' sex and the level of educational attainment of their father and mother.

The interview questions for the students were open-ended ones designed to probe deeper into the issues raised in the questionnaire. The teachers were asked open-ended questions about their own awareness of the issue, whether they mentioned the topic in their classes, whether they graded gender-exclusive or gender-inclusive differently, which form they thought students used most, and which form students were exposed to in their reading.

Data collection

Data were collected in three ways: the questionnaire completed by students, the interviews with students and teachers, and the students' writing. Students completed the questionnaire in a large lecture hall with guidance from the first author. They were told not to write their names on the questionnaire and assured that it was not connected to their grades. The first author then gave them a two-minute introduction to the topic, after which he went through each item, explaining it and giving students time to complete it before proceeding to the next item. After the questionnaires were completed, he gave students a 35-minute presentation on the larger issue of language variation and change. The presentation was done in order to thank the college for allowing the research to be conducted and to educate students on the issue. The presenter did not state his own views and practices on the issue - he supports and uses gender-inclusive usage but believes people should make their own informed decisions - until the end of the presentation, i.e., long after the questionnaires had been collected. After the presentation, interviews were conducted with the six students in groups of three in an open area on the campus. The interviews lasted about 15 minutes. The four teachers, none of whom had attended the presentation, were interviewed individually in a study room in the college library. Each interview lasted between 10-15 minutes.

As part of their preparation for the 'A' level exam, each student wrote a practice essay in response to one of 12 writing prompts similar to those which appear on the exam. This was done before the questionnaire data were collected. Afterward, writing scripts for three of the prompts were chosen for analysis, as these prompts seemed to generate many instances of third-person singular, an area in which writers often have to choose between gender exclusive/inclusive forms. The topics of these three prompts were: the meaning of love, the problems faced by teenagers, and whether or not it is worthwhile for a country to spend its time and money to train athletes for international competition. The typical script was approximately 633 words long, based on nine randomly selected scripts.

Data analysis

Descriptive statistics were obtained for the questions in section one of the questionnaire. Then, data on respondents' sex from section two of the questionnaire were used to run Chi-square tests to see if significant differences existed between females and males in their responses to items four and five of section one: whether they used gender-exclusive or gender-inclusive more frequently, and whether they believed gender-exclusive English is unfair to females, respectively. In item 4, responses indicating "I don't remember" were dropped from the chi-square calculation. A significance level of .05 was used for all inferential statistics in the study. Data from the interviews were combined with those from the questionnaire to better understand student behaviours in their essay writing, to which we now turn.

The writing scripts were coded into one of six categories based on the presence of gender-exclusive and gender-inclusive forms in the essays. The categories were: (1) gender-exclusive only; (2) gender-inclusive only; (3) mixed gender-exclusive and gender-inclusive, more gender-exclusive; (4) mixed, more gender-inclusive; (5) mixed, equal number of instances of gender-exclusive and gender-inclusive; (6) no explicit gender-exclusive or gender-inclusive. The coding was done by the first and second authors who first met to discuss randomly selected scripts. Inter-rater agreement, established by coding 8 scripts together and extensive discussion, was 100%. The first two authors then divided the rest of the scripts and coded them. While coding, any instances about which coding was questionable were noted for later discussion.

Rubin, Greene, and Schneider (1994) had coded the use of plural forms, e.g., "Architects should keep their clients informed" (Category 6 above) as instances of gender-inclusive. While the researchers in the current study believe that this is one of the best ways to implement gender-inclusive, it was not coded as such in this study because it was not possible to unambiguously see it in this way.

The resulting coding of students' writing on the practice exam was analyzed using descriptive statistics to see:
1. how many students were in each category; 2. how many used explicitly gender-inclusive forms (combining categories 2-5). To see if the writing of females and males differed in terms of the use of explicit gender-exclusive/inclusive forms, a chi-square test was conducted comparing the frequency of people of each sex in category 1 and those in a combination of categories 2-5.

To: Results

Back to G.A.L.E. Main Page This page last updated January 29, 2001

Web Manager
http://www.tokyoprogressive.org/gale/articles/asianenglishes/methods.html