One of the major challenges in developing a reading comprehension test is to select texts and tasks that are appropriate for test-takers at a given level of language proficiency. With the goal of helping test developers deal with such a challenge, this work investigates text complexity, test difficulty, and the relationship between these factors in a corpus of 25 renowned English reading comprehension tests. The investigation highlights synergy between qualitative and quantitative research approaches.

Castello organizes the discussion of his study into six chapters. The rationale and the theoretical and methodological framework are presented in the first chapter, followed by detailed descriptions of the specially compiled corpus of the reading comprehension tests in chapter 2. Subsequent discussion is based on analyses of text complexity (chapter 3). In particular, the complexity of input texts and accompanying test tasks is quantitatively examined through a number of linguistic and text measures (e.g., type-token ratio, lexical density, lexical variation, grammatical intricacy, and readability). Text complexity is further explored using systematic functional linguistics (e.g., semantics and textual cohesion) as a theoretical framework for qualitative analyses in chapter 4. The investigation of test difficulty was based on both quantitative and qualitative data obtained from test-takers’ reading performance, their perception of text and task difficulty, and actual test difficulty (chapter 5). Throughout the volume, relevant findings from quantitative and qualitative analyses are discussed and compared. The book concludes with an emphasis on the advantages of data obtained from diverse methodologies that can be triangulated throughout the analyses (chapter 6).

This thorough work makes contributions to the fields of corpus linguistics and language testing. One of the main contributions is that the results illustrate a way to examine reading texts and test tasks in terms of their complexity and difficulty. This work also demonstrates how quantitative and qualitative data can be used to supplement each other in the description of the relationships among text complexity, perceived test difficulty, and actual test difficulty.