VI.  METHOD

1. Hypotheses

 The present study tests the following two hypotheses:

1) The more reading that Korean-American parents do at home in their native
language, the better they serve as models of reading behavior for their children
(despite language differences). This in turn stimulates their children to do more
FVR in English.

2) The more FVR that Korean-American children do in English, the greater their
achievement in Academic English, as measured by their scores on the language
portion of the Stanford 9 Test.

2. Subjects

The subjects for this study were 99 American-born children and 99 Korean-American parents (for each child, the parent or other Korean-American adult closest to the child—typically the child¡¯s mother—was chosen as a subject). The criteria for inclusion in the study were as follows: 1) that all subjects reside in or near Glendale,8 California; 2) that all students currently attend 6th or 7th grade classes at a public school; and 3) that their parents subscribe to a Korean-language newspaper rather than to an English-language newspaper. Generally, Koreans who live in Glendale are middle class and well educated; in consequence, their socio-economic status is sufficiently controlled for the purposes of this study.

3. Materials

For measuring parental FVR, a list of 30 best-selling titles of modern Korean literature was compiled, to which 20 fictitious titles were added. For measuring the children¡¯s FVR, a list of titles of modern English-American literature was constructed by asking 50 6th grade honor students from California towns outside Glendale to list twenty of their favorite books. The titles of the 30 books occurring most frequently on those lists that typically are not recommended by the Department of Education were used in the test, along with 20 fictitious titles.9 This test was used to measure the amount of reading done outside of school. Title recognition required both students and parents to identify which titles are actual works of literature. For measuring the children¡¯s academic English achievement, an excerpt of the language portion of a Stanford 9 basic skills exam10 was selected.  

4. Procedure

Researchers studying reading behavior have found that when subjects are asked to estimate how much reading they do, they tend to exaggerate the amount (Beech, 1990). The title recognition tests used in this study not only avoid that defect but also can be answered quickly by subjects. 

1) The title recognition test for Korean literature was given to the mother of a Korean-American student except in cases where the student was not living with his mother; in those cases the test was given to whichever Korean-speaking adult was closest to the student, typically his father.

2) The title recognition test for English-American literature was given to 6th and 7th grade Korean-American students.

3) The score assigned to each test was computed by subtracting the percentage of fictitious titles checked from the percentage of genuine titles checked; in cases where this difference was a negative number, a score of zero was assigned.

4) The scores on (1) and (2) were analyzed for linear correlations to see whether the amount of parental FVR in Korean influenced the amount of their children¡¯s FVR in English.

5) An excerpt of the language portion of a Stanford 9 basic skills exam (hereafter English Stanford 9) was given to the 6th and 7th grade Korean-American students.

6) The scores on 2) and 5) were analyzed for linear correlations to see whether the amount of FVR by the students affected their academic English achievement as measured by their performance on English Stanford 9.