Re: Programmed instruction/EduBalony

An excellent history of Programmed Instruction can be found in a
paper by Vargas & Vargas (1992).  They cover the entire history
of PI and of teaching machines, extending the history back to various
pre-Skinner inventions and developments (e.g., early teaching and
testing machines such as Pressey's).

Programmed Instruction reached its zenith 1968-69.  You can chart
references and you can chart commercially available programs in print
(which Vargas & Vargas do).  Both indicators display PI reaching
a peak during 1968-69.  After that, the "bottom" fell out of the Programmed
Instruction movement.

Vargas & Vargas then cover various explanations for the demise of PI:

1. Inadequacy.  PI had become a bandwagon during the early to mid
1960s.   Many people got into the act.  A new society, the National
Society for Programmed Instruction, was formed.  Many commercial
enterprises were started.  And as a result, many bad programs
were turned out.  Much PI was "in name only," and Vargas & Vargas
note that these programs bore only a superficial resemblance to
the PI that Skinner and other experimental analysts of behavior had
developed.  Studies comparing much of this inadequate PI to
traditional forms of instruction found either no difference, or found in
some cases the traditional instruction to be better!

2. Hostility.  The Vargas's point out that the Educational Establishment
generally regarded anything behavioral with hostility.  PI gained its
biggest following in business and in the military.  It was a mixed
proposition in the schools.  Moreover, in the schools during the 60s
an interest in things scientific began to get replaced with a sandals
'n flower-child touchy-feely outlook to which PI was greatly incompatible!
Luddites entered the profession in droves!  They're still there, too,
by the way.

3. Ignorance.  The Vargas's point out that many of the people who
got on the bandwagon in the 60's had little or no background in the
requisite science.  They did not know either the experimental
analysis of behavior literature or its concepts and findings. Added
to that, for the PI movement there was no quality control agency
in place.  There was no equivalent to Underwriter's Labs or to
ANSI.  Anyone could break down material in to small chunks,
sequence it in some way and slap on a Programmed Instruction
label.

Both the producers and the consumers were ignorant of what PI
was supposed to be. They were ignorant of the science and its
derivative technology.  And they were ignorant of behavior and
behavior change.

4. Economics.  The Vargas's note that there was little profit to
be made in Programmed Instruction. There often was no economy
of scale.  Costs rose up to and sometimes exceeded benefits.
Companies, especially publishers, had a certain reluctance to
underwrite or sponsor either Research or Development in PI.

5. No Good Delivery Device.  Skinner and PI during the 50's and
60's lived and worked and developed this technology in the
pre-commonly-available Computer era.  The delivery devices for
PI were either paper-based, or mechanical via teaching machine.
Neither of these methods could respond well to student behavior.
The paper booklets and mechanical devices were incapable
of managing contingencies of reinforcement.

I will add a 6th reason:

6. Abandonment of Rate of Response.  This is a criticism that Og
Lindsley has often leveled against Programmed Instruction, and at
Skinner, and it's one such criticism that I share with Og.

Because PI was delivered by way of paper books and booklets,
or by mechanical devices, there was no method built into it
for tracking rate of response.  Developers of PI might get some
overall rates, such as how long it took a learner to complete
a disc, or to proceed through a set of frames.  But without
the cumulative recorder they could not see the rates well,
nor see any patterns or problems.  There was also no way for
them to view celeration (that being the pre-Chart era, too).

As a result of this measurement difference, perhaps, there was
no concept of fluency brought into PI.  PI might get a learner
to respond correctly and accurately, but without provision for
fluency, which in turn requires monitoring of frequency and
celeration, there was little or no attention to retention,
endurance, application, performance standards or to stability
(REAPS -- the effects of fluency).

Now, interestingly enough, standard measures and standard charts
might have provided the basis from which the developers of PI,
had they used these measures, had these charts, and had a
computer-based delivery platform, could have established that
missing quality control.  Had they had the quality control in place,
then this might have prevented the "in name only" versions of
PI from flooding the marketplace.  There'd still have been that
Education Establishment and its takeover by "60's People" to
have been dealt with, however.  -- JE

REFERENCE

Vargas, E.A., & Vargas, J.S. (1992). Programmed Instruction and
teaching machines.  In R.P. West &  L.A. Hamerlynck (Eds.)
Designs for Excellence in Education: The Legacy of B.F. Skinner.
Limited Edition. Longmont, CO: Sopris West.  pp. 33-69.