What makes an item difficult to read: insights on Spanish–English bilingual students’ decoding of Spanish words and pseudowords

This study investigated linguistic and orthographic factors affecting item difficulty parameters of items on Spanish word and pseudoword reading tasks, calibrated in a sample of Spanish-English bilingual first- and second-graders. For both words and pseudowords, we investigated the effect of word length, neighborhood size, and specific orthographic features (e.g., context-dependent graphemes, syllable blends, and diphthongs). We used adaptive least absolute shrinkage and selection operator (LASSO) regression to select the most relevant predictors, followed by a dominance analysis to evaluate the relative importance of the selected predictors in explaining item difficulty. We find that, for word reading, word length, context-dependent graphemes (c, g), and complex syllable structures like blends and diphthongs were associated with higher item difficulty, while neighborhood size was associated with reduced item difficulty. Similar results were observed for pseudoword reading, with length and syllabic complexity associated with higher difficulty and total neighborhood density and frequency with lower difficulty. We discuss additional task-specific patterns. These findings underscore the importance of considering linguistic and orthographic features when designing decoding assessments and instructional interventions for bilingual learners.