What is the difference between biserial and point biserial coefficient?

Dichotomous variables can be either a

discrete/true dichotomy or a continuous/artificial one. Examples of

discrete dichotomies are male vs female, dead vs alive. There is no

underlying continuum between the groups. Continuous or artificial

dichotomies are those which we assume there to be an underlying

continuum but we assign individuals to a category based on some

arbitrary criterion. Examples of this artificial dichotomy are pass

vs fail (based on some cutoff score on a test) or short vs tall

(based on some arbitrary height). A point-biserial and biserial

correlation is used to correlate a dichotomy with an interval

scaled variable. The difference is that the point-biserial

correlation is used when the dichotomous variable is a true or

discrete dichotomy and the biserial correlation is used with an

artificial dichotomy.