You are on page 1of 7

214 Q U A R T E R L Y JOURNAL O F EXPERIMENTAL PSYCHOLOGY

BROADBENT'S F I LT E R T H E O RY:
POSTULATE H AND T H E PROBLEM OF SWITCHING TIME
BY

N E V I L L E M O R AY
From the Department of Psychology, Hull University

Groups o f digits were presented b i n a u r a l l y a n d d i c h o t i c a l l y t o subjects w h o were


asked t o recall them. D i f f e r e n t rates and patterns of presentation were used, the design
being based o n t h a t o f Broadbent (1954). B r o a d b e n t ' s findings i n this field were con-
firmed. I f subjects are presented w i t h simultaneous pairs o f digits a t a rate o f 2 pairs
per second dichotically, t h e y cannot recall them i f t h e y are asked t o alternate between
the ears. I f however the presentation is staggered, so t h a t although the rate is constant
the signals do not overlap, subjects can recall alternately from the two ears. I t is suggested
t h a t these findings a r e b e t t e r interpreted as a n interference effect, n o t a r a t e effect.
Criticism i s offered o f s i m i l a r designs t o measure " s w i t c h i n g r a t e . " C r i t i c i s m i s also
offered of Broadbent's estimate of "perception t i m e " i n such experiments. A n analysis
of the q u a n t i t y and type of errors made b y subjects is given, which suggests t h a t Broad-
bent's theory of a short term store on the peripheral side of a selective filter is in need of
revision.
INTRODUCTION
Broadbent (1954) presented pairs o f digits simultaneously and dichotically t o
subjects. T h e signals came in groups of three pairs, and subjects were required to
recall the digits either successively ( L L L R R R ) o r b y alternating between the ears
(LRLRLR). B r o a d b e n t found that the latter was only possible providing that the
rate of presentation was not greater than one pair every i sec., although if recall were
successive the subjects could recall accurately even when the rate was as high as
2 pairs per sec. H e interpreted these results t o mean t h a t a t t h e faster rates
the subject could not switch from ear to ear fast enough, and that the only successful
way to deal with the situation was to listen to one ear while the material arriving at
the other ear was retained in a short term store on the peripheral side of what he
called a "Filter Mechanism."
Now it could be argued that the above experiment is not a fair test of a subject's
ability to switch between signals : for what in fact a subject is doing in such a situation
is to switch between the signal he hears and a trace of the other, not the other signal
itself. I t seemed that i t should give a more genuine measurement of the subject's
performance i n a switching task i f the signals came not i n simultaneous pairs, b u t M O I L .
staggered, so that they never actually overlapped. T h e following experiment was
devised.
EXPERIMENT I
Method
Lists of six digits were prepared, and presented t o the subjects i n groups of ten lists.
There is a r o sec. pause between each group of six during which the subject had to recall.
A Brennell M k . I V two-channel tape recorder w a s used t o present t h e s t i m u l i t o t h e
subjects t h r o u g h headphones, w h i c h w e r e B r o w n m o v i n g coil, l o w impedance t y p e .
Subjects were asked t o recall the digits either in any order they liked (free recall), o r all
the digits f r o m one ear followed b y a l l those f r o m t h e o t h e r (successive recall), o r b y
switching from ear to ear (alternate recall). A binaural condition was also used, in which
cases t h e s t i m u l i were obtained b y passing t h e above dichotic lists t h r o u g h a m i x i n g
box and then feeding t h e m to b o t h headphones. T h e free recall conditions were given
before any of the other conditions so t h a t all subjects were well practised. T h e dichotic
conditions of presentation were as follows: simultaneous ) a t z signals/ear/sec.;
staggered a t 2 signals/ear/sec., and slow overlapping ( ) at

You might also like