Statistics for HCI. Alan Dix

       Preface

      Sometimes, it seems we are bombarded with numbers, from global warming to utility bills. In user research or academic studies, we may also encounter more formal statistics such as significance testing (all those p-values) or Bayesian methods; and graphs and tables, of course, are everywhere.

      For those of us working with people, we know that numbers do not capture the complexities of social activity or the nuances of human feelings, which are often more appropriately explored through rich qualitative studies. Indeed, many researchers shun anything numerical as, at best, simplistic and, at worst, dehumanising.

      However, the truth is that we all use statistics, both in our work and day-to-day lives. This may be obvious if you read an article with explicit statistics, but mostly the statistics we use are informal and implicit. If you eyeball a graph, table of results, or simple summary of survey responses, and it affects your opinions, you are making a statistical inference. If you interview a selection of people or conduct a user trial of new software and notice that most people mention a particular issue or have a particular problem, you are using statistics.

      Below the surface, our brains constantly average and weigh odds and we may be subconsciously aware of statistical patterns in the world well before we explicitly recognise them. Statistics are everywhere and, consciously or unconsciously, we are all statisticians. The core question is how well we understand this.

      This book is intended to fill

Скачать книгу