Page 151 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2015 Vol. 41

150 Journal of Library Science in China, Vol. 7, 2015

Qingpu, respectively. Thirdly, one junior middle school and one senior middle school were
randomly selected from each of the five districts mentioned above, and from the area within and
out of inner ring in Pudong New District as well. With the above three steps, a list of 14 schools (7
junior middle schools plus 7 senior middle schools) was established. Fourthly, 14 undergraduate
students were recruited from the Shanghai Jiao Tong University and trained as investigators.
One investigator was in charge of the delivery of the questionnaires in each school. The study
focused on students of Grade One and Grade Two, excluding Grade Three in which students
were busy preparing the coming entrance examination to high schools or colleges at that time.
The investigator firstly selected one class from Grade One, and delivered the questionnaire in
the classroom during the break with the permission of the teacher. The students who had not any
experience of Reading via WeChat were filtered by enquiry before the questionnaires were handed
out. Then the investigator went to the class with the same number in Grade Two, and repeated the
same procedures. He/she delivered and collected the questionnaires among students of Grade One
and Grade Two in an alternative way, until 100 copies were collected in each school.
After two weeks’ hard work, the present study successfully collected 1 400 questionnaires. After
dropping the cases with partial answers, we finally had a valid sample of 1 039, none of which
had any missing data. The sample of 1 039 students comprised 470 (45.2%) boys and 569 (54.8%)
girls; there were 506 (48.7%) junior middle school students and 533 (51.3%) senior middle school
students. Regarding the area where they are living, 599 (57.7%) students are from urban area while
440 (42.3%) are from suburb area.

3 Data analysis and results

3.1 Exploratory factor analysis of the social reading motivation

The major goal of this study was to identify the dimensions of the social reading motivation as a
construct. The original motivation scale contained 22 items. This analyses yielded a KMO value
2
of 0.947, χ =15 084（df=231，Sig.<0.001), which was considered reasonable to conduct a factor
analysis on the matrix. The transformed data for the 1 039 samples was subjected to a principal
components analysis with varimax rotation, and six factors were yielded finally. The items measuring
the social purpose of reading adapted from MRQ by Wigfield and Guthire (1997) and those
measuring social purpose of social media use adapted from the scale by Park, Kee, and Valenzuela
(2009) loaded on one factor. Additionally, one item (“Reading via WeChat can help me forget worries
temporarily”) measuring the variable of Killing Time loaded together with the other four items that
measured the variable of Self Development, however, the loading value was less than 0.5.
Realizing the fact that the item’s loading value on the new factor was rather low, this study
deleted it finally. Table 1 presents the result of the factor analysis after deleting the item. As it
shows, the study still yielded six factors, which accounted for 75.6% percent of the total variance.