Wolfgang Viechtbauer

Marginally significant (p = .07)

User Tools

Site Tools


profile

My (Not So Really) Short Profile

Most people probably don't read more than the first paragraph anyway, so I'll give you the short (and dry) version first. The not so really short details follow after that.

I am associate professor of methodology and statistics in the Department of Psychiatry and Neuropsychology (Faculty of Health, Medicine, and Life Sciences) and the School for Mental Health and Neuroscience (Division 2: Mental Health) at Maastricht University in the Netherlands. I got my Ph.D. at the University of Illinois, Urbana-Champaign. I do research on meta-analysis, mixed-effects models, and other fun stats things and provide methodological/statistical support for my colleagues who do research on the causes and treatment of psychiatric disorders.

I was born and raised in Aachen, Germany, a city of about 250,000 peope located near the border to the Netherlands and Belgium (there is in fact a tripoint of all three countries just outside of the city, at the Vaalserberg, where you can happily jump from one country to the next – plus, the Vaalserberg, clocking in at a whooping 322 m (1,058 ft), is the highest point in the Netherlands, so as you can imagine, it is a sight to behold!). At the quaint age of 16, I spent a year as a foreign high-school student in the United States, during which I decided to stay in the US for my university studies. During the next 10 years, I first studied psychology at Illinois State University (which got me interested in academia and research), but then decided to switch my focus to statistics / psychometrics (which is what I am really passionate about). Fun fact: One can also use statistics to show that a PhD in statistics is the best grad degree you can get!

I received M.S. and M.A. degrees in statistics and psychology from the University of Illinois, Urbana-Champaign, where I also completed my Ph.D. in 2004, specializing in statistical methods as used in the social, behavioral, and health sciences. I then moved back to Europe (Bush got reelected – kinda left me no choice) and started working at Maastricht University in the Netherlands, first in the Department of Methodology and Statistics and now in the Department of Psychiatry and Neuropsychology.

As a statistician, my research involves the comparison of existing and the development of new methods for the analysis of data generated in various fields of research. I am primarily interested in the methodological aspects of systematic reviews and the statistical methods used for meta-analyses, but my interests more generally encompass the design and analysis of longitudinal and multilevel studies using appropriate random-effects and mixed-effects models. If that sounds like a bunch of gobbledygook to you, just remember that data science is the sexiest job of the 21st century – rawr!

Probably the most useful contribution of my academic career – at least the one that I think had the most impact – has been the creation of the metafor package, a free and open-source add-on for conducting meta-analyses with the statistical software environment R. A paper published in the Journal of Statistical Software (Viechtbauer, 2010a) describing the package has been cited over 10,000 times (which is likely to reflect only a fraction of its actual use). More information about the package and its ongoing development can be found on the metafor package website.

In addition, I have written some more technical papers over the years, comparing and developing statistical methods for analyzing meta-analytic data with random/mixed-effects (meta-regression) models (e.g., Jackson et al., 2014; Lopez-Lopez et al., 2014; Viechtbauer, 2005, 2007a, 2007b, 2007c, 2010b). Along the way, I have also provided help and assistance on a number of applied meta-analyses and primary research projects/studies (including a number of RCTs) pertaining to a wide variety of topics (e.g., personality, cognition, physical therapy, cardiovascular disease, psychopathology, medication adherence, smoking cessation, diabetes management, school violence, childhood abuse, psychotherapy). Here and there, I also may have had something potentially useful to say about other methodological issues and methods (e.g., de Bruin, 2012; Kotz et al., 2012; Viechtbauer, 2007d; 2010c). If you would like to find out more, you could take a look at a list of my publications (citation info on Google Scholar – not like I care about things like that).

Within the Department of Psychiatry and Neuropsychology, I provide statistical/methodological support for my colleagues and try to provide living proof that statisticians can be pretty friendly and socially adapted individuals after all (p = .07, but this definitely suggests a trend bordering on statistical significance). The research conducted within the department is focused on the mechanisms through which interactions between individuals and their environments contribute to the development of major psychiatric disorders and somatic comorbidity.

What does that mean? Well, think about it this way: A pair of monozygotic twins is genetically identical (well, mostly, kind of), yet it can easily happen that one twin develops a particular disease while the other one does not. This may suggest that environmental factors are at play. Or think about a group of people living in the same area, sharing a roughly similar economic status, family structure, diet, and so on, yet some individuals develop a particular (noninfectious) disease while others do not. This may suggest that genetic factors are at play. Now put these two things together and you get questions such as: What genes, or more specifically, what biological pathways are involved in the processes that determine, to some degree, why some people are more affected by a stressful environment than others? This question asks about a gene-environment interaction and this is a core part of the research conducted within the department. The goal is to figure out why some people develop things like major depressive disorder, anxiety disorder, bipolar disorder, schizophrenia, or some general form of psychosis, while others do not, and what can be done to help afflicted individuals gain insight into their condition and hopefully overcome it.

Things like stress sensitivity are probably not well measured by just asking a bunch of individuals how much stress they have ("So, how much stress do you have?") and how they feel ("How is it going?") – and then checking if those who say "Lots" also tend to say "Totally bad" and vice-versa. You don't need to bring in a statistician to figure out that part (or if that's about the most sophisticated advice you can get from your local statistician, somebody made a bad hiring decision …). Instead, a lot of the research in the department involves the so-called experience sampling method, where individuals are prompted to respond to a questionnaire (including questions about affect and various contextual variables) multiple times throughout the day, sometimes for up to a week or even longer.

So, we collect lots of repeated measurements about things like stress and how a person feels at various points throughout the day and then examine the data for patterns and relationships (e.g., explore variables that may determine why some individuals react more strongly to stressful events than others) and how such differences (which are often rather subtle) can put some individuals in a more vulnerable state for developing certain conditions than others. This is where things get interesting and the statistical analyses get complicated. And that's when they call me in: The Statistician (cue the dramatic music and slowly fade in the opening credits).

Well, thanks for reading (assuming you didn't skip all other paragraphs except the first). Feel free to get in touch at wvb (at) wvbauer (dot) com – sorry, this keeps the spam down a bit – especially if you want to chat about fun things like the use of informative priors on the heterogeneity variance component in meta-analytic models, generalized linear mixed-effects models with mixture distributions for the random effects, and the Wiener distribution (lol!). When I am not geeking out on stats, I like to spend time with my wife and our two sons, do some gardening, play the piano and chess, try my hand at woodworking, and occasionally fly radio-controlled gliders/sailplanes (which I rarely crash – honestly!).

profile.txt · Last modified: 2024/03/17 19:02 by Wolfgang Viechtbauer