This survey was approved by the research committee of the School of the Humanities, National Southeast University, China. The survey was administered between February 2011 and October 2011 and involved one university in Nanjing and one university in Chongqing.
A cohort of 1200 Chinese college students randomly received the test battery. The participants were selected from all of the college students in which students take part in an English course (3647 students) in the universities, and one in every three the participants were selected according to their students identification number. Of the 3647 students, 1215 students were selected to receive the survey, and among them, 15 students refused to participation. Of the 1200 test batteries, 88 were not returned and 64 were incomplete and thus excluded, for a response rate of 92.5 %. Of the participants, 508 were females (48.4 % of respondents) and 540 were males (51.6 % of respondents). The mean age of the female students was M = 20.6 years (SD = 1.2) and that of the male students was M = 20.8 years (SD = 1.1). School officials, lecturers and students had previously received a written description of the survey. The students were also informed that participation in the study was voluntary and that they could withdraw at any time. The students responded to the self-administered questionnaires in a classroom during a 20- to 30-min session.
The living conditions were measured by a trained researcher who physically visited one of the dormitory rooms and determined the dimensions using a tape measure. The size of the rooms in Chongqing and Nanjing was standardized by gender. In Chongqing, four female students shared an 19.48 m2 by 2.90 m high room, while four male students shared an 18.65 m2 by 2.90 m high room. In Nanjing, four to six female or male students shared a 23.64 m2 by 3.41 m high room.
All the participants completed the questionnaire, which included demographic questions such as the age of the participant and parental education as well as a battery of three tests, namely Zung’s Self-Rating Anxiety Scale, the Internet Addiction Test, and a Self-Control Scale, and questions about binge eating and compensatory behaviors.
The demographic characteristics, which included ethnic group, gender, age, weight, ideal weight and body height were self-reported by the participants.
Zung’s Self-Rating Anxiety Scale
This questionnaire (Zung 1971) is a 20-item, 4-point self-reported measurement designed to measure symptoms of anxiety. The scale, which has been translated into Chinese and previously administered in China, demonstrated good validity and reliability (Wu 1990, 1999; Tao and Gao 1994). All 20 items on the survey had been previously adjusted to measure anxiety as perceived to be caused by population-density. For example, the original first item, “Do you feel more nervous and anxious than usual?” was adjusted to read “when there are a number of people in your dormitory room, do you feel more nervous and anxious than usual?” Similar adjustments were made to the remaining 19 items. In the present survey, the Cronbach’s alpha for female and male participants was 0.81.
Internet Addiction Test
This questionnaire (Young 1998) is a 20-item, 5-point self-reporting measurement designed to assess the degree of respondents’ Internet use. Higher scores reflect a higher level of problematic Internet use. The questionnaire exhibits adequate reliability and validity (Widyanto and McMurran 2004). In the present survey, the Cronbach’s alpha for female participants was 0.85 and it was 0.89 for male participants.
The Internet Addiction Test was translated into Mandarin Chinese by the first author and was administered to Chinese participants with acceptable validity and reliability (Tao 2013).
The Self-Control Schedule
The questionnaire was developed by Rosenbaum (1980) and contains 36 items that enable the participants to rate their level of self control (Sinha and Nayyar 2000).
The SCS has been reported to be a reliable measure, and a number of studies have supported its validity as a measure of self-control behaviors (Rosenbaum 1980). In the present survey, the Cronbach’s alpha was 0.88 for female participants and was 0.82 for male participants.
The SCS was translated into Mandarin Chinese by the first author. The Chinese version was then retranslated into English by an independent translator (an associate professor of English). The retranslated version was found to closely match the original and any differences between the two versions were reconciled.
Binge eating and compensatory behaviors
At the beginning of the study, we explained it to participants the symptoms of binge eating as according to the DSM-IV criteria (APA 2000) i.e., recurrent episodes of binge eating in which the amount of consumed food is significantly higher than that consumed by most people in similar conditions. Binge eating is defined as the consumption of a large amount of food (e.g. a large carton of ice cream or a large bag of chocolates) in a short period of time (e.g. 2 h), accompanied by a loss of control over the eating behavior, that is followed by feelings of guilt. There were two questions on the survey that assessed the respondents’ binge eating characteristics. The first question was “Have you engaged in binge eating in the last 3 months?” The response options were “yes” and “no”. The second question was “How frequently have you engaged in binge eating in the last 3 months?” The response options included “one or more times a day”, “two to six times per week”, “once a week or less” and “no binge eating”. The questions have been used in a previous study (Tao 2013).
Recurrent inappropriate compensatory behaviors
Recurrent inappropriate compensatory behaviors such as (vomiting, excessive exercise and diet pills) are one of the DSM-IV criteria for bulimia nervosa (APA 2000). Two questions were used to determine the respondents’ tendencies to engage in these behaviors. The first question was “Have you engaged in compensatory behaviors (such as self-induced vomiting, excessive exercise, or taking diet pills) to prevent weight gain during the last 3 months?” The respondents answered either “yes” or “no”. The second question was “How frequently have you engaged in these compensatory behaviors to prevent weight gain during the last 3 months?” The response options included “one or more times every day”, “two to six times per week”, “once a week or less” and “no compensatory behaviors”. The question has been used in a study (Tao 2013).
Analyses were performed in SPSS 18.0 (SPSS Inc., Chicago, IL, USA), and a multivariable regression was used to identify the significant predictors of binge eating, compensatory behaviors and Internet addiction.
Interaction terms were included in the analyses to determine whether binge eating, compensatory behaviors and anxiety were moderated by self-control. Interaction terms were also included in the analyses to determine whether Internet addiction and anxiety caused by high-density living condition were moderated by self-control.
We also addressed multicollinearity by centering the continuous variables. The VIF scores, which varied from 1.00 to 2.14, suggested that multicollinearity was not a substantive problem in the data, thus allowing for a meaningful interpretation of the results.
Statistical significance was based on two-sided tests evaluated at a 0.05 level of significance.