The term “behavioral sciences is vastly too broad to cover in one sweep. The issue of statistics and how the math is used in the behavioral sciences is as varied as the sciences, themselves are!
First, Psychology was the only recognizable behavioral science. Now, there are hundreds, if not thousands of fields that fall under the category, which has expanded as far as to create such new fields as Physiopsychology which has now split off into Neuroscience. Neuroscience incorporates areas that are not part of the traditional Psychological field, so the statistics used these days are completely out of the realm of the traditional behaviorist.
How decisions are made is now the sub field of “Decision Design”. Cognitive science has to tackle issues that were not even present decades ago. Behavioral genetics incorporates previously unheard of scientific areas.
For example, there is a multitude of models which have been developed over the years: including computational models which include algorithms, protocols and network mapping. And this is just to manage the information processing aspects of managing the computations and the huge amounts of data required to investigate such areas as “how do people make decisions?”
There are statistical models, most of which determine the relationships between pairs of variables, the way that observations form in groups or deviate from groups, or can be forecast to exhibit different behaviors.
In statistical models, the basic statistical unit in behavioral science is the person. Then there are observations of something about the person, perhaps once, perhaps over time. There is information in the way that the observations behave or exhibit information.
The basic statistical unit may be a group of people, or a population. Then there are observations that can be examined to see if they provide information that allows prediction or probability work.
In other areas, sub units of the statistical unit, such as the women in a population, or an organ in the human being become the focus. then there are models for examining the behavior of organs under various forms of stimulation and determining whether probability can be extracted.
The list goes on and on. And the lists, complexities, and details grow more and more every day. Even surveys, such as mail surveys, in person interviews, online surveys, and surveys that are conducted over the cell phone can take a lifetime of study and development, as the results are analyzed in order to predict the behavior of small, medium or large populations.
In other words, it is not Grandma’s world of behavioral sciences anymore, and statistics is not Grandma’s world of statistics, either.