How hard should the books be in elementary small-group reading? This study explored text difficulty for bilingual students reading below grade level in third grade. Using a convergent parallel mixed methods design, I used qualitative methods to analyze students’ engagement and discussion during small groups and single case design to evaluate students’ fluency and reading comprehension after reading and discussing texts in small groups.

Six Spanish-English bilingual students, split into two groups of three, participated in twelve, 30-minute, small-group reading sessions. Students in Group 1 read approximately one year below grade level, and students in Group 2 read approximately a year and a half below grade level. In six of the twelve sessions, students read and discussed texts matched to their reading levels, and in the other six they read and discussed texts one year ahead of their reading levels. I assigned matched and difficult texts across the twelve days by blocked randomization.

I analyzed video transcripts of each session to understand students’ engagement (focus of engagement, strategies, and interaction) and discussion (inferential vs. literal responses, instances of verbal participation). At the end of each session, students reread and retold the book the group had read and discussed that day to produce a fluency (words correct per minute) and comprehension (ideas correctly retold) score.

Findings were complex and revealed that different levels of texts have both advantages and drawbacks. Key findings included: For fluency, half of the students benefited from matched texts. The other half read difficult texts with similar fluency to matched texts. For comprehension, text difficulty did not matter for anyone except one student, and for him it only had an effect on 3 of 12 days. Group 2 engaged much more with texts and ideas in difficult books and with pictures in matched books. Group 1 had more inferential/interpretive responses with matched texts, and Group 2 had more inferential and interpretive responses with difficult texts. Most students participated evenly regardless of the difficulty of the text under discussion. However, two students talked more when discussing matched texts.