I recently reviewed a learning chapter for a gen. psych.
textbook and asked the same question. The book basically
says: "Punishment is bad", but doesn't offer any references.
The strength of reinforcement is measured by number of
responses and resistance to extinction. So I guess punishment
schedules should be judged by two similar criteria: Number
of responses and persistence(?) of extinction.
A quick search in PsycInfo is pretty interesting.
923 hits for "partial and reinforcement and schedules"
(101 since 1990)
52 hits for "partial and punishment and schedules"
(5 since 1990)
Here's an example of the latter which suggests that the answer is
more complicated than "punishment is bad."
Effectiveness of fixed ratio punishment and durability of its effects.
ABSTRACT Hooded rats pressing a bar for food reward on a fixed interval
schedule were shocked after every 1, 2, 4, or 8 responses with .6-ma shocks
of .25-, .50-, or 1-sec duration. All Ss eventually ceased responding. In
retest sessions 2 wk. later, several Ss who had received the least frequent
punishment were among those who did not resume bar pressing. Only for the
shorter shock durations was partial punishment less effective than
punishment for every response. ((c) 1999 APA/PsycINFO, all rights reserved)
AUTHOR Storms, Lowell H.; Boroczi, George
AFFILIATION U. California, Los Angeles
SOURCE Psychonomic Science. 1966 5(11) 447-448
But my specialty is not learning, so I could be missing something
obvious. Hope there are other responses.
TT
===================================
Thomas A. Timmerman, Ph.D.
Assistant Professor
Psychology Department
Austin Peay State University
Clarksville, TN 37044
931-221-1248
[EMAIL PROTECTED]
===================================
