Research
Research Interests
My research program has been about making improvements to the
scientific methods used in the social and behavioral sciences, in
an effort to produce a more accurate and cumulative literature. My
research evaluates, improves, and develops research methods of a
statistical and measurement nature for the fields that use
psychological, behavioral, or social data (e.g., psychology,
sociology, management, marketing, education, behavioral medicine).
My research focuses on the methods of designing studies and
analyzing data for the social and behavioral sciences, which are
arguably the most foundational aspects of an empirical science.
Without an appropriate design or if impoverished analysis methods
are used, the value of the research is questionable and leads to a
literature filled with suspect conclusions, thereby limiting the
effectiveness of what should be a living and cumulative
literature. My efforts in this space have helped to reduce various
methodological shortcomings. A general way of saying what I do is
work on methods of designing studies and analyzing data.
Additionally, I apply a variety of methods collaboratively with
others in mutually beneficial collaborations in a variety of
domain specific areas, where I can develop needed or apply
existing methods to address interesting and important real-world
problems. As Tukey pointed out, “the best thing about being a statistician is that you get to
play in everyone’s backyard!”
Primary Area of Research-General
My primary research is on the interrelated topics of effect sizes,
confidence intervals, and sample size planning. Sample size
planning is one of the most important aspects of designing an
empirical study, because using a sample size that is much too
large for the particular research goal potentially puts more
participants than necessary at risk, delays dissemination of
findings, and is not an effective use of limited resources. Using
a sample size that is too small for the goal, however, lowers the
likelihood that the research goal can be addressed with enough
confidence to add to the literature or ensure that the
participants’ and researcher’s time was used wisely. The design
stage of an empirical study is where researchers can have arguably
the biggest impact on success, and where potential methodological
shortcomings can be prevented (instead of attempting to fix
later).
General Research Interests
My interests span widely across the field of research methodology
– I just do not have enough time to work on all of them with the
same intensity as I do for research design! Some of the other
topics that I work on are longitudinal data analysis,
mixed-effects models/multilevel models, mediation models, general
latent variable models, finite mixture modeling, statistical
classification and statistical discrimination, the bootstrap
technique, the proper design and implementation of Monte Carlo
simulation studies, and various psychometric issues. The methods
that I am interested in need not be conceptualized as being
mutually exclusive, as many times the methods are combined to form
a unified approach to designing research studies and analyzing
data. Further, much of what I do involves statistical computing
and R is involved in much of my work. An interest related to all
others is the cross-fertilization of methods from a variety of
fields. Methodological developments in one field are often not
well known in other fields, even though both fields may ask
questions that can be addressed with the same or similar methods.
By working in a variety of fields in an interdisciplinary fashion
and borrowing methods from each, better methodological practice
can be implemented in each field, which is beneficial all around.
General Research Interests
My interests span widely across the field of research methodology
– I just do not have enough time to work on all of them with the
same intensity as I do for research design! Some of the other
topics that I work on are longitudinal data analysis, general
latent variable models, finite mixture modeling, statistical
classification and statistical discrimination, the bootstrap
technique, the proper design and implementation of Monte Carlo
simulation studies, and various psychometric issues. The methods
that I am interested in need not be conceptualized as being
mutually exclusive, as many times the methods are combined to form
a unified approach to designing research studies and analyzing
data. An interest related to all others is the cross-fertilization
of methods from a variety of fields. Methodological developments
in one field are often not well known in other fields, even though
both fields ask questions that can be addressed with the same or
similar methods. By working in a variety of fields and borrowing
methods from each, better methodological practice can be
implemented in each field and all fields benefit.
Overarching Research Goal
The overall goal of my research is to evaluate, improve, and
develop research methods of a statistical and measurement nature
for fields that use psychological, behavioral, or social data
(e.g., management, marketing, education, behavioral medicine,
information technology, sociology, psychology) so that substantive
questions can be addressed with quality methods.
Collaboration
I have been a consultant on many research projects ranging from
small scale narrowly focused studies to large scale government
funded projects. Feel free to contact me if you think my research
could be beneficial to your research. Depending on many factors, I
may or may not be able to provide assistance and/or collaborate.
Get Involved
Graduate students interested in getting involved with
methodological should feel free to
contact me.