Students
Reference: Sánchez L.Z. (2004). Effects of Parent Participation Using First Language Curriculum
Materials on the English Reading Achievement and Second-Language Acquisition of Hispanic Students. Unpublished PhD Thesis. Lehigh University, Pennsylvania, USA.
Research Question
“Will using the students' curriculum materials, translated into their native language (i.e., Spanish), as an at-home paired reading intervention improve their English reading achievement at differing ESOL levels (i.e., level 1 versus level 2)?” (p18)
Participants
Eight 4th and 5th grade (ages 9-11) children enrolled in regular classes at an inner-city elementary school in Allentown, Pennsylvania, USA, in a district characterised by the author as approximately 45% Hispanic. Participants were born either in Puerto-Rico or Dominican Republic and had been in the USA for between three months and three years. Selection to participate was based on three characteristics: their ESOL (English for Speakers of Other Languages) teacher’s assessment that they were “having difficulties and considered at-risk in reading as they acquired a second language” (p57); a moderate or high level of cognitive academic language proficiency (see Cummins 1980) in Spanish, their L1, as determined by the Woodcock-Muñoz Language Survey; and their parents’ capacity to deliver the intervention at home. The eight participants who were selected were split evenly between ESOL level I (new to English) and ESOL Level II (limited English proficiency). One participant dropped out of the study before receiving the intervention, leaving seven participants whose data were analysed.
Interventions and comparisons
The parents of participating children delivered the experimental intervention as follows. Each week a set of reading materials slated for use at school in the following week were translated into Spanish and sent home. Parents were trained in a pedagogic approach known as Paired Reading. Parents were then asked to use this approach to read and discuss the materials with their children at home two to three times over the course of the week, for 15 to 20 minutes at a time. A multiple baseline, interrupted time series design was used to allow a comparison between slope values of outcome
scores in the ’pre-intervention’ period with slope values of outcome scores in the ‘with-intervention’ period.
Outcomes
The principal outcome measure was the participants’ oral reading rate in English - the number of words in a passage of text they could read out loud correctly in one minute. Slope values were calculated by comparing the scores from one week to the next to describe the development of each child’s oral reading rate. Measurements were taken twice on each occasion, once using the original English-language versions of the translated texts, and once using age appropriate English-language texts which the participants had not seen before, in either Spanish or English. The latter assessment, called a ‘generalization’ score, was used to determine whether the effects of an intervention that focused on specific texts could be said to transfer to reading proficiency more generally. In addition to oral reading rate, participants were assessed on their comprehension of the texts, the
development slopes of which were also calculated.
Finally, average scores for each child on the two outcome measures in the ‘pre-intervention’ period were compared with average scores in the ‘with-intervention’ period.
Results
In 19 of the 28 comparisons of ‘pre-intervention’ and ‘with-intervention’ slope values, the slope value decreased after the intervention had been implemented.
Mean scores in the ‘with-intervention’ period were higher than in the ‘pre-intervention’ period.
Authors’ conclusions
Sánchez concludes that the results “do not support the hypotheses that a paired-reading
intervention implemented at home in the native language (i.e., Spanish) using classroom materials would significantly improve the students English reading achievement” (p95).
Comment
Sánchez’s primary conclusion is sound, based on the data she has presented. However, later in her discussion, she states that “[a]lthough significant results in ORR and comprehension were not observed based on slope metric results after implementation of the paired-reading intervention, good level effects (i.e., mean differences), as reflected by effect sizes (greater than 1.0), were
obtained by most students suggesting that there was substantial growth over time for participants” (p96). This optimistic interpretation of the results has been informed by a comparison of the mean scores for each child in the periods before introducing the intervention and in the period during which the intervention was being conducted. Far from providing any meaningful assessment of the effects of the intervention, however, Sánchez is simply stating that the participants’ reading
improved over time. Moreover, given that slope values tended to decrease after the introduction of the intervention, that is to say the intervention appears to be associated with a negative effect on the outcomes of interest, it is interesting to speculate on what the character of the overall
improvement in reading might have been had the intervention not been implemented at all. In the absence of a control group we are not in a position to say.
Table 3.7 Risk of bias assessment for Study 2
Selection Bias Study Design Confounders Blinding Data Collection Method Withdrawals and Dropout Moderate: Participants selected by the researcher Moderate: Within subject differences over time. High: No randomisation Moderate: Not reported High: Validity not described Low: 1 participant (12.5%) dropped out