In 2010, I was seconded from my classroom teaching position to work as a facilitator in the implementation of the newly-established National Standards. In this role, I ran in-depth Literacy and Assessment workshops in schools. I found that despite the national effort to close the gap between achievers and non- achievers, schools are still faced with the issue of most Samoan students not doing well as they should be, especially in the area of literacy. Assessment procedures from which such data is derived are widely researched in New Zealand and abroad. In spite of the numerous studies conducted on literacy assessment, none has ever been done to critically analyse the linguistic nature of these literacy assessment tools; yet these tools and their implementation may well contribute to the phenomenon of underachievement among Samoan students in New Zealand. Thus the aim of this inquiry was to open up the practice of standardised reading tests to examination. This thesis examines unacknowledged issues in the design of the standardised reading tests by which students are assessed in multicultural contexts. A Samoan critical approach was adopted specifically for this research: ‘Tofā'a'anolasi’, which draws on Foucault’s analytical tool box to counter-read tests texts and practices, including the way achievement data is managed and applied. Tofā'a'anolasi is based on my interpretation of Foucault’s ideas about power, knowledge and language in society, and how these ideas are useful in educational research undertaken from a Samoan perspective. This research framework is used to examine assumptions within and about assessment practice, particularly as it pertains to Samoan students in New Zealand. This thesis shows how the normalising hegemonic nature of assessment has profoundly disadvantaged Samoan students. This study includes critical discourse analysis of selected standardised reading testing tools in New Zealand primary schools. For this work, two reading test papers routinely used throughout New Zealand to assess the reading knowledge and skills of Year 7 and 8 students were analysed and critiqued. I also gathered stories from teachers as assessment administrators and Samoan students as assessment takers. Through focus group conversations, these assessment participants shared their own understandings and critiques of assessments. Moreover, I communicated with three key informants, who are also experts in the area of New Zealand primary school literacy assessments about the two standardised tests examined in thesis. Their stories are included in this research thesis. The analysis of the findings reveals serious mismatches between the language, culture, knowledge, experience and interests most of the Samoan students bring into the testing context, and those expected to succeed in the standardised tests. These discrepancies cause students confusion and dismay. The analysis also reveals the predominantly male oriented nature of tests contents and administration. In this thesis, I argue that the male orientation of tests affects some Samoan students in two ways. First, the assumption of male superiority (and female inferiority) is in direct contrast with the Samoan culture of ‘feagaiga’, or the sacred covenant and deep respect Samoan brothers hold for their sisters. Second, the male and female themed nature of tests contents rely on a male/female binary within tests which tends to exclude third gender, for instance, the Samoan fa'afafine. In this thesis, the three emerging themes of linguistic bias, cultural bias and gender bias are presented with examples from the tests examined. The review of the literature reveals that standardised assessment is a governmental practice; its processes are constructed within historical and cultural notions of what is normal and what is not. This categorisation bolsters the unquestioned power of assessment to regulate its subjects as governable individuals. The power of assessment causes particular behaviours, such as teachers teaching the content knowledge to suit the purposes of the state for the students to learn. The contribution of this study lies in highlighting issues that are taken for granted in the design and practice of standardised reading tests in New Zealand primary schools. These issues include sometimes the inescapable biases in the standardised reading test papers, in terms of how their language, content and presentations of gender and ethnicity work against a marginalised student population such as Samoans living in New Zealand.