Elementary Schoolers ’ Attitudes toward Reading in English : How Boys Feel Relative to Girls

This paper describes the results of a study that examined the attitudes of Malaysian primary school boys and girls toward reading in English as a Second Language. Using the Students’ Reading Attitude Survey, which the researchers adapted from McKenna and Kear (1990) Elementary Reading Attitude Survey, 2,666 responses were analysed using the non-parametric Mann-Whitney U test. The findings reveal that the students had positive attitudes toward recreational and academic reading in English as a second language, but that the girls scored significantly higher on all the dimensions of reading attitudes compared to the boys. The results provide some insights into the gender gap between boys and girls in reading literacy.


Reading Achievement and Reading Attitudes
Research has shown that reading achievement is related to positive attitudes toward reading (Baker & Wigfield, 1999;McKenna et al., 1995;Walberg & Tsai, 1985).Wigfield and Guthrie (1997), for example, found that motivated students read more books and spend longer time reading, which is expected to bring about better performance in reading; Russ and Mark (1988) who surveyed the attitudes of 7 th graders, and Druyor (2012) discovered that students who have positive attitudes toward reading also tend to have high achievement in reading; while Wade (2012) found the same correlation between positive attitudes toward reading and achievement in the areas of math, science, and social studies.Furthermore, evidence also indicates that having a positive attitude toward reading will most likely lead to sustained, lifelong reading (Brown, 2001;Cannon, 1997;Logan & Johnston, 2009;Pandian & Ab. Latiff Ibrahim, 1999).

Reading Attitudes and Gender
Research also indicates that students' attitudes toward reading vary according to gender (Baker & Wigfield, 1999;McKenna, 1997;McKenna & Kear, 1990).A study by McKenna (1997) on 269 students found that the majority regarded reading as a predominately feminine activity.They were of the opinion that reading is an activity more suitable for girls and that those perceptions intensified with age.Clark and Burke (2012) conducted a longitudinal study in the United Kingdom and found that girls outperform boys on all National Curriculum reading tests; girls enjoy reading more and do it more often and have more positive attitudes toward it; and they seek out more reading opportunities (e.g.library visits) compared to boys, who tend to read less often, and think less positively about reading.Research conducted by Chatterji (2006) also shows that boys have consistently lagged behind girls in reading; from when they were enrolled in kindergarten until the end of first grade where the gap is the widest.In a broad cross-sectional study of 14,315 students, McQuillan (2013) found that the girls had more positive attitudes toward reading than the boys.
Numerous studies conducted in the second (L2) or foreign language (FL) setting have also found differences in the attitudes of girls and boys toward reading.Azlina Murad Sani and Zaizati Zain (2011), who investigated adolescents' attitudes toward reading in a setting where English is not highly demanded outside the classroom found a lack of healthy attitudes toward L2 reading among the boys.The girls, on the other hand, held a significantly more positive attitude toward L2 reading in general.The same findings were found by Butler (2007), who carried out a study on children's reading attitudes in their L1 and an FL, and Gallagher-Brett (2006), who discovered that the overall attitudes toward reading in their L2 were also more positive among the female respondents.
Given the strong evidence supporting the fact that students' attitudes toward reading could influence their success in school and beyond (Black, 2006;Coles & Hall, 2002;McKenna, Kear, & Ellsworth, 1995;McKenna & Kear, 1990;Schmitt, 2009), and how these attitudes vary according to gender (Baker & Wigfield, 1999;McKenna & Kear, 1990), it is of vital importance that this issue be studied to narrow the gender gap in reading.

Research Questions and Hypotheses
This paper presents the results of a study that examined primary school students' reading attitudes in their L2.In particular, the study investigates the following research question: What are the students' attitudes toward recreational and academic reading in English?Do boys and girls differ in terms of their attitudes?
The following hypotheses were tested to address the research question: Hypothesis 1 H 0 : There is no difference in the attitudes of boys and girls toward recreational reading in English.
H 1 : Girls have a more positive recreational reading attitude in English than boys.

Hypothesis 2
H 0 : There is no difference in the attitudes of boys and girls toward the academic reading in English.
H 1 : Girls have a more positive academic reading attitude in English than boys.

Instrument
To gauge the boys' and girls' attitudes toward reading in the L2, the researchers used the Students' Reading Attitude Survey (SRAS), adapted from the Elementary Reading Attitude Survey, or ERAS (McKenna & Kear, 1990), which has been used extensively in studies looking at students' attitudes toward reading (e.g.Butler, 2007;Worrell, Roth, & Gabelko, 2007) because of its estimates of a high reliability, and evidence of construct validity, which were based on a national sample of more than 18,000 school children, grades 1-6 (McKenna et al., 1995;McKenna & Kear, 1990).The high reliablity of the instrument is estimated from the values of Cronbach's alpha (between .74 to .89),measuring the internal consistency of the two subscales of academic and recreational reading (McKenna et al., 1995), while the evidence of construct validity was obtained in different ways for each subscale.The recreational subscale was trialed by categorising children into two groups: a) those with and without library cards (given that they had access to a library) and b) those with and without a book checked out from the school library (given that there was no requirement for them to check out books).Evidence for the academic subscale was obtained through the categorisation of children based on reading ability.
For the purpose of the study, the researchers made some minor amendments to the language and response icons of the original instrument to ensure that it fits the Malaysian context, but still maintained the essence and spirit of the original questionnaire.For example, some American expressions were replaced with Malaysian ones to ensure that the students clearly understood the questions.These include replacing phrases such as, 'on a rainy Saturday' with, 'on a weekend'; 'summer vacation' with, 'the long school holidays'; 'bookstore' with 'bookshop'; and 'reading class' with 'English language classes'.In addition, instead of using the cartoon and comic strip character, Garfield, for the different levels of responses, the researchers substituted them with smiley faces to denote emotions ranging from, 'very happy', 'quite happy', 'not very happy', and 'not at all happy' to avoid any confusion and uncertainty for those Malaysian children who may not be familiar with the Garfield character.
The ERAS by McKenna and Kear (1990) is tailored for English speakers.Hence, to help the Malaysian students-whose native language is not English-to answer the questions regarding their attitudes towards reading in English, the researchers modified the questions by adding "in English" to all the twenty original statements.This was to point out to them that the context was about their reading attitudes in the L2.The researchers prepared the survey in two languages-Malay and English, to ensure that the questions can be understood by those students that were not conversant in English.Because the Malay language is the official or national language of Malaysia, the students are more likely to understand it than they would English.

Selection of Sample
The random cluster sampling technique was employed to select the sample because it is the most appropriate for and best suited to the school context.Random cluster sampling involves "the random selection of naturally occurring groups or units (clusters)….[such as] universities, schools, school districts, classrooms, city blocks, and households" (McMillan, 2008, p. 117).Before data collection, all the schools in the state of Selangor, Malaysia were listed, and then six schools were randomly selected from the list.All the Year 1, Year 2, and Year 3 students took part in the study; thus allowing the researchers to obtain a response rate of 100%.Table 1 shows the number of students that participated.

The Choice of Statistical Test
The Mann-Whitney U Test was employed to test for the differences in reading attitudes between boys and girls as this test is the most suitable test for the given data and assumptions made about the population.The normality test conducted on the data had provided strong evidence to conclude that the scores in the population are not normally distributed as the review of the Kolmogorov-Smirnov (K-S) test for normality (K-S = .106,df = 1343, p < .05),skewness (-1.098), and Kurtosis (1.746) statistics suggested that the assumption of normality was violated.Given the importance of the assumption of normality, among others, for the justifiable use of parametric tests (Brewer, 1996;Cohen, 1977;Seigel & Castellan, 1988), and the serious consequences on the validity of the probability statements made under the given assumptions (Brewer, 1996;Glass, Peckham, & Sanders, 1972) the non-parametric Mann-Whitney U test was chosen, as it does not make any assumptions about normality given that the analyses are conducted on ranked scores (Field, 2005;Green, Salkind, & Akey, 2000;Milenovic, 2011;Pallant, 2007;Seigel & Castellan, 1988).

Adequate Sample Size
In addition to fulfilling the requirements of random selection and proper choice of statistical test for a given set of assumptions, the researchers also conducted a power analysis (See Cohen, 1977, Statistical Power Analysis for the Behavioural Sciences) to determine the size of the sample that is needed to allow for a power (the probability of correctly rejecting the Null hypothesis) of .99;an  (the probability of making Type I error, or the probability of incorrectly rejecting the Null hypothesis) of .01;and an effect size (the size of the effect that the researchers wanted to capture before they would consider the results to be important) of .20.This is to allow for the effect size to be detected with the highest probability of being correct (power of .99)and lowest probability of being incorrect ( of .01).The sample size needed was 1,084 (see Brewer, 1996;Cohen, 1977;Mohd-Asraf & Brewer, 2004;Seigel & Castellan, 1988).Hence, the sample size of 2,666 that was obtained was sufficient to justify the values of α, power, and ES that the researchers had set for the purpose of hypothesis testing.

Recreational Reading Attitude
Descriptive statistics show that girls (median = 35, mean rank = 1480) scored higher on recreational reading attitude than boys (median = 32, mean rank = 1156) for reading attitudes in English.The Mann-Whitney U value was found to be statistically significant, U = 651202 (z = -10.98),and p < 0.01.Therefore, the null hypothesis was rejected in favour of the alternative hypothesis that girls have more positive recreational reading attitude compared to boys.
The distribution in scores on the recreational reading attitude measure in English for girls is higher than the boys as shown in Figure 1.

Academic Reading Attitude
The Mann-Whitney U test also revealed a significant difference in the academic reading attitude levels of boys (median = 32, n = 1334) and girls (median = 34, n = 1296), U = 698146, z = -8.56,p = .00,r = -0.17.The distribution of scores on the academic reading attitude measure for girls is also higher than the boys as shown in Figure 2. Hence, the null hypothesis that there is no difference between boys' and girls' academic reading attitude in English was also rejected, in favour of the alternative hypothesis that girls have a more positive academic reading attitude compared to boys.As shown in Figures 1 and 2, the medians of boys' and girls' recreational and academic attitudes are quite high, which demonstrates their positive attitudes toward reading.However, although both boys and girls showed positive reading attitudes, it was found that the girls scored significantly higher on all the dimensions of reading attitudes compared to boys.The findings corroborate the earlier findings of McKenna, Kear, and Ellsworth (1995), who also found that girls held more positive attitudes toward both recreational and academic reading than boys.It is also consistent with other studies that found that females have better reading attitudes than males in the L2 (Azlina & Zaizati, 2011;Bas, 2012;Huang, Liang, & Chiu, 2013;Uusen & Müürsepp, 2012).

Discussion and Conclusion
The results show that both the recreational and academic reading attitudes of girls are more positive than that of boys, a pattern that has consistently been noted in previous research (e.g.Guthrie & Wigfield, 2000;McKenna et al., 1995).This indicates the need for reading instruction in L2 classrooms to take into consideration their differences in attitudes as to foster the development of students' reading interests in the L2.Hence, teachers should design high-quality instructional practices that correspond to the needs of students (Wiggs, 2012) and that focus on developing boys' (and girls') positive reading attitudes from the start of schooling, as this study has shown that the variations in reading attitudes between boys and girls are present even in the early primary school years.
The results also suggest that teachers may benefit from employing a repertoire of teaching strategies in the reading classrooms.The mismatch between teachers' teaching strategies and students' learning needs and abilities can impact upon their learning as they may not address their academic reading needs.Consequently, students' attitudes toward reading, as well as their level of engagement, can be affected.
Since teachers play an important role in encouraging students to have positive attitudes about reading, Gettys and Fowler (1996) in Herron-McCoy (2009) have suggested that teachers who enjoy reading, and who give pleasurable reading opportunities and allow for recreational reading during class time, can promote a positive reading culture in the students.A positive reading habit beyond academic reading should be encouraged, as "academic reading can be developed in a way to promote recreational reading outside the classroom" (Herron-McCoy, 2009, p. 53).
To help teachers address the fact that boys and girls may have different needs, relevant parties, for example, the school administration, should provide some assistance in the form of training, knowledge-sharing sessions, and up-to-date information regarding gender differences so that teachers are aware of, and consequently work out possible ways to address the issue.By empowering teachers, the students will be able to be helped to grow and reach their potential beyond the classrooms to be life-long readers.Finally, the teachers and the school should also create the awareness among the parents of the need for differentiated learning, as well as request for their assistance so as to inculcate positive reading attitudes at home to ensure that there is a positive extension and continuation between what happens at school and at home.

Figure 1 .
Figure 1.Distributions of the scores for English recreational reading attitude

Figure 2 .
Figure 2. Distributions of the scores for English academic reading attitude

Table 1 .
Number of participants being administered the Student Reading Attitude Survey