This study investigates the challenges associated with statistically evaluating the Simple View of Reading model (SVR) and its applicability to the Swedish language. The SVR model, a widely-accepted and popular framework, posits that reading comprehension (RC) is a product of two independent factors: decoding (D) and language comprehension (LC), expressed as RC = D x LC. While various statistical approaches have been employed to validate the model in English, a deep and non-transparent orthography, a consensus on a formal testing method has not been reached. Additionally, the model's functionality in other languages, such as Swedish, which has a semi-transparent orthography, remains unclear.

 This study has two primary objectives: 1) to longitudinally test the SVR model's validity from year 1 to 3, examining the relative contributions of decoding and language comprehension factors to reading comprehension over time using a latent variable approach, and 2) to assess the model's applicability to the Swedish language. To achieve these goals, we utilize an extensive dataset from the LegiLexi foundation, comprising data from 43,127 students across 2,666 schools and 18,006 classes.

 The presentation will discuss methodological and statistical considerations necessary for evaluating the SVR model, as well as the contributions of decoding and language comprehension factors to reading comprehension during primary school's learning-to-read process.