This chapter explains why we think that statistical methodology mat-ters so much to the HCI community, and why we should attempt to improve it. It introduces some flaws in the well-accepted methodology of Null Hypothesis Significance Testing, and briefly introduces alterna-tive. Throughout the book we aim to critically evaluate current prac-tice in HCI and support a less rigid, procedural view of statistics in fa-vour of “fair statistical communication”. Each chapter provides scholars and practitioners with the methods and tools to do so.
|Title of host publication||Modern Statistical Methods in HCI|
|Editors||Judy Robertson, Maurits Kaptein|
|Place of Publication||Switzerland|
|Number of pages||14|
|ISBN (Electronic)||ISBN 978-3-319-26633-6|
|ISBN (Print)||ISBN 978-3-319-26631-2|
|Publication status||Published - 4 Apr 2016|