Many words in English resemble each other in multiple ways. When these words have similar spelling, they are referred to as orthographic neighbors. The purpose of this within-subject experimental study was to examine the effect of orthographic neighbors on the acquisition of spelling, more specifically constructing orthographic representations of words. Five questions will be addressed in the study: (1) Is there an effect of orthographic neighbors on the acquisition of orthographic representations? (2 Is there an effect of phoneme-to-grapheme consistency on spelling acquisition? (3) Is there an effect of delay on spelling acquisition? (4) Can participants’ ability to learn spelling improve without specific instruction? (5) Are there interactions between the effects of orthographic processing ability and the effects of rime, substitution, and transposition neighbors on participants’ spelling acquisition?

Seventy-one second grade students in northern California participated in the study. Following assessment of participants’ cognitive ability, five sessions of a computer-based experiment were conducted. In each session, the participants were shown two base words and attempted to learn seven new words. The seven new words included control words (without orthographic neighbors: no connection with the corresponding base word) and target words (representing various types of connections with the corresponding base word). Each new word was presented three times for five seconds in random order. Orthographic choice and spelling tasks were used to assess orthographic learning via the orthographic choice and spelling tasks. Each task was administered twice: immediately after the experiment and two days after the experiment.

Three-level logistics regression with random effects for sessions and participants were used to analyze the data. The model allows between-student variation in learning outcomes due to the individual differences among participants and contextual effects at the session level. The outcome variable was participants’ performance on the orthographic choice and spelling tasks. Level 1 included the words’ characteristics, Level 2 the sessions’ characteristics, and Level 3 the participants’ characteristics. In addition, random coefficients of effects of neighbors (Rime, Substitution, Transposition) and the cross-level interaction between orthographic neighbors and the participants’ orthographic processing ability were considered.

The effect of rime neighbors on participants’ performance were found in both the orthographic choice and the spelling tasks. The effects of substitution and transposition neighbors existed only for participants’ performance on the orthographic choice tasks. The facilitative effect of phoneme-to-grapheme consistency was found in the orthographic choice tasks, but not in the spelling tasks. The effect of delay was found both in the orthographic choice tasks and the spelling tasks. Orthographic processing was as a significant predictor for participants’ performance on all four posttests. A significant interaction between orthographic processing and rime neighbors was observed in both spelling tasks. The findings demonstrate that second-grade students can use orthographic analogies to facilitate their orthographic learning and that orthographic processing is an important cognitive ability for spelling acquisition.