<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body text="#008080" bgcolor="#FFFFFF">
<b>*Apologies for double posting* </b><br>
<br>
<br>
Dear colleagues, <br>
<br>
You are cordially invited to a presentation by Sean Wallis (Survey
of English Usage, UCL), in which he will talk about his book
Statistics in Corpus Linguistics Research: a New Approach/
*(Routledge 2021). <br>
<br>
The presentation will be on <b>Thursday 7 April 2022, 15:00</b> and
will take place online. A link will be sent around a few days before
the event. <br>
<br>
Below you will find the abstract of the talk, as well as some
information about the book. <br>
<br>
Best wishes, <br>
<br>
Evelien Keizer (University of Vienna) & Gunther Kaltenböck
(University of Graz) <br>
<br>
<br>
<b>Book Launch: /Statistics in Corpus Linguistics Research/
(Routledge 2021) - Sean Wallis, Survey of English Usage</b><b><br>
</b><b> </b><b><br>
</b><b>
Abstract:</b><br>
<br>
Why do people find 'statistics' difficult, and what can we do about
this? What are the best methods to use in linguistics, and are there
specific problems we must address when we apply statistical methods
to corpora? <br>
<br>
In his new book, Sean Wallis argues there are several reasons why we
find statistical reasoning counter-intuitive. Probably the most
fundamental is that we do not "see" sampling uncertainty, we have to
count many events, which is often an impossible task. But with a
computer we can calculate and visualise uncertainty on the same
scale as an observed factor, which is what /confidence intervals/
do. Whereas traditional approaches to confidence intervals were
inconsistent with statistical testing and sometimes obtained
improbable events, modern methods do not suffer these defects, and
may be extended into a wide range of testing environments. <br>
<br>
Applying these methods to corpus linguistics requires us to address
a number of challenges and traditions. For example, conventionally,
many statistical approaches accepted linguistic variables with per
(million) word baselines. Yet these are clearly suboptimal, as most
phenomena can only occur in specific locations in a text. This is
fundamentally a linguistic analysis problem, which must be addressed
through good research design, well-considered queries and a careful
review of data. <br>
<br>
Other problems tackled in the book include questions of
semasiological analysis, learning how to engage in statistical
argument to reduce research workload and how to compensate for the
fact that corpora are random samples of texts, rather than random
samples of independent utterances, clauses or phrases. <br>
<br>
<b>From the jacket:</b><br>
<br>
Traditional approaches to statistics focused on significance tests
have often been difficult for linguistics researchers to visualise.
/Statistics in Corpus Linguistics Research: A New Approach/ breaks
these significance tests down for researchers in corpus linguistics
and linguistic analysis, promoting a visual approach to
understanding the performance of tests with real data, and
demonstrating how to derive new intervals and tests. <br>
<br>
Accessibly written for those with little to no mathematical or
statistical background, this book explains the mathematical
fundamentals of simple significance tests by relating them to
confidence intervals. With sample datasets and easy- to- read
visuals, this book focuses on practical issues, such as how to: <br>
• pose research questions in terms of choice and constraint; <br>
• employ confidence intervals correctly (including in graph plots);
<br>
• select optimal significance tests (and what results mean); <br>
• measure the size of the effect of one variable on another; <br>
• estimate the similarity of distribution patterns; and <br>
• evaluate whether the results of two experiments significantly
differ. <br>
<br>
Appropriate for anyone from the student just beginning their career
to the seasoned researcher, this book is both a practical overview
and valuable resource. <br>
<br>
<br>
<pre class="moz-signature" cols="72">--
Univ.-Prof. Dr. Evelien Keizer
Institut für Anglistik und Amerikanistik / Department of English
Universität Wien Campus d. Universität Wien
Spitalgasse 2-4/Hof 8.3
1090 Wien
Austria
Homepage: <a class="moz-txt-link-freetext" href="https://anglistik.univie.ac.at/staff/staff/keizer/">https://anglistik.univie.ac.at/staff/staff/keizer/</a></pre>
</body>
</html>