Event Phone: 1-610-715-0115

Cancellation Policy: If you cancel your registration at least two weeks before the course is scheduled to begin, you are entitled to a full refund (minus a processing fee of $50).

In the unlikely event that Statistical Horizons LLC must cancel a seminar, we will do our best to inform you as soon as possible of the cancellation. You would then have the option of receiving a full refund of the seminar fee or a credit towards another seminar. In no event shall Statistical Horizons LLC be liable for any incidental or consequential damages that you may incur because of the cancellation.

A Distinguished Speaker Series Seminar by Benjamin Bolker, Ph.D.

Statistical models that account for grouping or clustering in data (variously called multilevel, mixed, or hierarchical models) have been around since the 1950s but have continually grown in scope and flexibility. Any time we have observations that violate the standard assumption of independence–because multiple observations are drawn from the same individual, classroom, species, or geographic region–our statistical models must incorporate this dependence structure or else risk inaccuracy and overconfidence.

While the computational tools for fitting and evaluating these models are ever more powerful and convenient, that power brings problems as well. When the data are too sparse for a complex model, how should we simplify or adjust the model? How do we choose among a bewildering range of options? How well do these complex models scale to big data, however defined?

The first hour of this seminar reviews the ideas of mixed models, discusses available software platforms, and briefly outlines some of the challenges in estimation and inference for these models. The second hour presents a general workflow for mixed modeling, illustrated with one or more examples. The third hour revisits the challenges from the first hour in more detail and discusses open questions. Examples will primarily be drawn from ecology and psychology, but the techniques are broadly applicable. Each hour will include a mix of examples, methods, and discussion, including responding to your questions.

The course is not hands-on, but it will include analyses using R.

Attendees will benefit most from the seminar if they have some knowledge (even dimly remembered) of basic statistics (e.g., normal distribution, hypothesis testing) and linear regression. Some knowledge of R will make it easier to follow the examples.