There are a large number of Hispanic students in schools today that are learning English, with Spanish as their native language. Some school districts have Bilingual education programs, as did the district in the Carlisle and Beeman’s (2000) study.
Denton, Anthony, Parker and Hasbrouk, (2004) also investigated strategies that work in bilingual programs. They attempted to discover whether the Read Well or Read Naturally programs helped bilingual students in three different areas: word identification in lists, word attack (phonemic decoding), and passage comprehension.
The participants in the study were 93 students ranging from the 2nd to the 5th grade in five different schools in one Texas district (Denton, et al., 2004). All the students were learning English as a second language (ESL). All the participants in the study were Hispanic and spoke Spanish as their first language. Participants in the study included 48 males and 45 females. The students were in bilingual classrooms. The school used a transitional bilingual program, so in second grade the teachers used mostly Spanish as the language of instruction, and by the time the children reached 5th grade, the language of instruction was predominantly English, with instruction gradually shifting to English over the elementary years.
Students’ pretest scores on the word attack subtest of the Woodcock Reading Mastery Test- Revised placed them into one of two groups (Denton, et al., 2004). These groups were emergent decoding and established decoding in English. The students were placed in matched pairs based on these test results. There was an attempt to have the pairs be from the same classroom if possible. 1 comparison/control group for Read Well and 1 for Read Naturally groups consisted of the matched pairs of students. There were 2 experimental groups that studied with either Read Well or Read Naturally. These programs were not compared with each other, but with their own comparison group. Due to attrition, nineteen of the students who finished the study were in the Read Well treatment group and fourteen were in the Read Well comparison group. Thirty-two were in the Read Naturally treatment and twenty eight were in the Read Naturally comparison.
The students in the treatment groups received tutoring with undergraduate university students using either the Read Well or Read Naturally program. They were tutored three times per week for 40 minutes each session over a ten week period. There was some individual tutoring, and some students were in small groups of two, three, or four.
Groups were formed based on scheduling constraints within the school.
The Read Well program was found to help with students’ decoding, but not with comprehension in comparison to the control group (Denton, et al., 2004). The tutoring focused on the parts of pronunciation that are different in Spanish and English, so the students could use their prior knowledge of the Spanish language to learn how the languages differed. Students who were tutored with Read Well gained 4.06 points on average in decoding. Only context free reading was improved. Word identification was the only factor that showed statistically significant gains (f=5.70, p=.023). Word attack showed a mean gain of 5.16, though the comparison group gained 2.35, and in comprehension thee mean gain for the treatment group was 1.58, only .01 more than the control group. After completing tutoring with this program, students were able to read English words in that they could pronounce them fluently, but they could not understand what they meant.
Using the Read Naturally program, there was no significant difference from the control group on any of the criteria. Word attack had a mean gain of -.22, and word identification had a mean gain of only 1.12 (comparison group gained 1.75, more than the treatment). The highest gain for this group was only a 2.13 point improvement for the treatment group on passage comprehension (the control group also improved .71 of a point on the same measure).
One problem with the internal validity of this study is that the way groups were chosen was not elaborated beyond the constraint of scheduling within the schools.
Students groups varied greatly; from six students who received one on one tutoring to groups of two, three, and four students. I would say that a student who is tutored for 40 minutes one on one will have a much greater amount of progress than a student who receives 40 minutes of instruction in a group of four. The tutor can not attend to the child’s individual needs as much in a larger group. Also, there is no information given as to the tutoring group size for the two experimental treatments. Did Read Well and Read Naturally treatments have similar student to teacher ratios? Additionally, there was no information given on what students were missing in class to participate in the treatment. Were they missing the literacy block in their classroom? This could have had an effect on why, in some instances, the control groups surpassed the treatment groups in their mean gain in scores on the Woodcock Reading Mastery Tests.
De la Colina, Parker, Hasbrouck, and Lara-Alecio, (2001) also explored the use of Read Naturally on students learning to read in Spanish. Their subjects were in a school district in Texas that provided bilingual instruction for students in kindergarten through fifth grade, in an attempt to transition them into English speaking and reading. This is done out of the belief that students learn to read in a second language more easily if they are fluent in their first language.
A review of the literature showed the researchers (De la Colina et al., 2001) that aspects that improved English speaking children’s ability to read with automaticity were rereading the same passages, modeling by a teacher, and the students’ ability to monitor their own progress.
Materials from the Read Naturally program were translated for the students (De la Colina et al., 2001). The program entails students repeatedly reading the same passage, listening to tapes of the passage and reading along, and then notifying the teacher when they are ready to test.
Subjects for the study included first and second graders from four different classrooms (De la Colina et al., 2001). All students chosen had to be able to read between 30 and 60 words per minute in Spanish, or be able to read between 50 and 100 sight words to qualify for the study.
Students were split into three groups, one received instruction for a 12 week period, one received instruction for 10, and the last received instruction for only eight weeks (De la Colina et al., 2001). Each group met three days per week for 45 minutes. Small groups, mixed by engagement level as well as classroom assignment (to control for teacher affects) were formed to receive the intervention. One problem with this design may be that the groups were staggered so the first group started two weeks before the second, and in another two weeks the third group started. The researchers admit that students in the later groups may have gotten competitive and worked harder in order to catch up to the growth of their peers who started first. A way to fix this problem could be to simply start all the groups at the same time, and stop instruction for the shorter term groups earlier. Instead of starting the eight week treatment on week four, start them on week one and terminate their treatment at week eight.
Students’ engagement level was determined by the number of stories they read each week (De la Colina et al., 2001). This was not a reflection of reading level, because lower level stories were simpler and shorter, so they could be read faster. Highly engaged readers read a mean of 5.8 readings per week, and low level readers read only a mean of 2.6. These groups were both comprised of both high and low reading level groups.
Results of the study found that highly engaged students improved much more than lower engaged students, regardless of the amount of weeks they were tutored for (De la Colina et al., 2001). Two of the lowly engaged groups performed worse as the intervention went on. For the 12 week intervention, low engagement students did not have any statistically relevant improvement. The improvements that were made were somewhat modest, and could have been due to the regular classroom teaching and were similar to progress that may have occurred, even without the intervention (De la Colina et al., 2001). Ten of the 12 groups did improve over the course of the study, and those who were highly engaged improved between two and four times as much as students who were less engaged.
I find that this study (De la Colina et al., 2001) to be transferable to the extent that a student’s engagement, regardless of language, may affect their progress. I think the results are not as credible as they could be. The design was flawed by the effects of students working hard to catch up to their classmates, and the fact that the length of time intervention was received provided little difference is not explained. It is possible that floor effects in the ability of the program to help students improve caused the similarity of results among groups, and this should have been attended to.
Riccio et al. (2001) also worked with students learning to read in Spanish. They recognized the importance of phonological awareness in beginning reading acquisition, and the goal of the study was to investigate correlations between phonological awareness and Spanish and English reading ability.
Participants for the study (Riccio et al., 2001) attended three different Texas elementary schools. The school district had a bilingual education program. One hundred and forty nine participants were recruited, including 71 girls and 78 boys. All of these children were classified by their parents as Hispanic. The ages of subjects ranged from 5 to 11 years, they attended kindergarten through fifth grade. The majority (126) of the subjects were in bilingual classrooms, while 6 children moved back and forth between bilingual and English only classrooms during their day, and 17 were in classrooms conducted entirely in English.
Examiners who administered the tests to students were all bilingual (Riccio et al., 2001). Four measures were found to be especially important, and subtests from the Comprehensive Test of Phonological Processes (CTOPP) were used to measure English phonological awareness in initial sound matching, ending sound matching, rhyming, and deletion. The Conciencia Fonologica en Espanol (CFE) was created by the researchers and reviewed by a panel of experts in the bilingual education field and bilingual Hispanics from different cultural groups. This test also measured the ability to distinguish and match initial and ending sounds, identify words that do and do not rhyme, and the ability to delete phonemes. Students were also asked to read a short passage for one minute in each language. The Spanish version came from Read Naturally, and a doctoral student/translator translated it into English. The report did not clarify if the students received a different passage or the same translated passage in each language.
Having the same passage would have affected the results, because whichever one the students read first, they would be familiar with the content, even if the second time it was in another language. The background knowledge is there the second time and it is easier to guess the word.