Mixed designs make use of already-present variables and manipulate a second variable. This is also referred to as a quasi-experimental or natural design. Subjects are not randomly assigned to groups; they automatically fall into one of those categories.
Mixed designs are used when a result is further distinguished by another independent variable. For example, how fast a person runs is also delineated by age, gender and race.
A mixed design in psychology is one that contains both within- and between-subjects variables. One such design provided by Psychology World is called a pre-post-control design where the subjects are their own controls and are compared using two types of therapy.