ORDERING The paperback version of this book is sold out and will not bereprinted.
An eBoek version became available november 2015. the eBook costs € 9,- (US $ 10.-) — Available via Kobo/iBooks (Apple)/Kindle(Amazon.nl) — Or take a look at the sample chapterfor free.
Advising on research methods: Selected topics 2014 results from a research master course Methodological Advice that was given at the
University of Amsterdam at the end of 2014 by Gideon J. Mellenbergh and Herman
J. Adèr.
The objectives of the course were: (a) to acquire methodological knowledge that is
needed for advising researchers in the behavioral and social sciences, and (b) to get
experience with methodological consultancy.
The main material for the course was the book:
Advising on research methods: A consultant’s companion by Herman J. Adèr and Gideon J. Mellenbergh (with contributions by David J.
Hand). See: ARM book.
The students had to fulfill various assignments, one of which was to write a paper on
a topic that may come up during methodological consultancy.
In the beginning of the course, paper topics were selected from a long list of relevant
methodological issues.
The publication process of drafting, submitting, reviewing, adapting and
correcting was the same as in the production of any other edited paper
collection.
Eleven students participated in the course. They made the following contributions
to the book:
Eline J. S. S. Tan and Bianca Westhoff
discuss the use of unobtrusive measurements in behavioral
and social science research. Unobtrusive measurements are procedures
where participants are unaware of being measured. Four types
of unobtrusive measurements are distinguished: simple observations,
contrived observations, physical traces, and archives. The authors
describe the advantages, disadvantages, and ethical aspects of
unobtrusive measurements. The advantage of unobtrusive measurements
is that they are not vulnerable to respondent biases, such as, social
desirable answering questions. Disadvantages are, for example, that the
reliability and validity of unobtrusive measurements are often hard to
assess, and that ethical constraints may occur. The authors recommend
to supplement obtrusive measurements with unobtrusive ones if that is
possible.
Boris Stapel and Ruben J. J. van Beek
discuss the influence of a third
variable on the relation between two other variables. The authors
distinguish three types of third variables: confounders, moderators, and mediators. They describe the status
of each of these three variables. Moreover, they discuss statistical
methods to test moderator and mediator effects. Finally, they bring
together questions that are often asked on confounders, moderators, and
mediators, and give short answers to these questions.
Yoram K. Kunkels and Pia Tio
give an introduction to network analysis to visualize the structure of psychological concepts. The article starts
with an introduction to network analysis. Subsequently, network analysis
is demonstrated by an example: An analysis of test takers’ responses to
items that measure the Big Five personality traits. Moreover, the article
mentions the association measures and software to apply network analysis
in practice.
Maria C. (Riëtte) Olthof
addresses the analysis of clustered data, for example, achievement test scores
of students within school classes. Clustering of data violates the
independence assumption that is often made in statistical methods. She
describes the strong negative effects that clustering can have on the Type
I errors of statistical tests. She discusses simple methods to correct for
the effects of clustering. Moreover, she gives an impression of multilevel
analysis for the analysis of clustered data. She compares methods to
analyze independent data with simple and multilevel methods to analyze
clustered data using language test scores of students nested within classes.
The example shows large differences between conclusions drawn from
the analysis of clustered data with and without taking into account the
clustered nature of the data.
Noor Seydel and Martina van Slooten
address the topic of equivalence null hypothesis testing. Equivalence
null hypothesis testing applies to the situation where researchers want to
ascertain that variable means of different groups of study participants are
approximately the same. They discuss the construction of an equivalence
interval, which is the interval where researchers consider the difference
between two means negligible. Moreover, the authors describe how an
equivalence null hypothesis is tested. The article ends with a tutorial
that demonstrates how equivalence null hypothesis can be applied in
behavioral research.
Quentin F. Gronau and Tom H. Oreel
introduce readers into Bayesian inference and the inclusion of prior information into Bayesian analysis. The
article starts with a short introduction into Bayesian concepts and
methods, and the inclusion of prior information. The article shows the
steps to update prior information, and methods to study hypotheses.
The methods are demonstrated by data of studies on suicide attempts
and deaths. The example shows that a Bayesian analysis can yield other
conclusions than a conventional (non-Bayesian) analysis.
CONTENTS
Preface About the authors Unobtrusive measurements by Eline J. S. S. Tan and Bianca Westhoff Confounders, moderators and mediators by Boris Stapel and Ruben J. J.
van Beek Network representation as a visualization tool by Yoram K. Kunkels and
Pia Tio The analysis of clustered data by Maria C. Olthof Equivalence null hypothesis testing by Noor Seijdel and Martina van
Slooten Including prior information in Bayesian procedures by Quentin F. Gronau and Tom H. Oreel References