00100	THE OBJECTIVE MEASUREMENT OF AUTISTIC CHILDRENS' INTERACTION
00200	   PATTERNS WITH A COMPUTER-CONTROLLED PROGRAM FOR
00300	   THE STIMULATION OF LANGUAGE DEVELOPMENT
00400	
00500	
00600	            KENNETH MARK COLBY
00700			  AND
00800		    HELENA C. KRAEMER
00900	
01000		One  difficulty  in  evaluating  treatment  improvements   in
01100	disorders  whose  defining  characteristics  are purely behavioral is
01200	that we lack objective measurements. An objective measurement is  one
01300	which is intersubjectively confirmable and impartially independent of
01400	individual opinions, intuitions, and judgemments.  The  fallibilities
01500	of  humans  trying  to be both participants in and observer-recorders
01600	of,  their  interactions  are  such  that  objective  measurement  is
01700	impossible.   But when one participant in an interaction is a machine
01800	such as a computer, an opportunity arises for the machine  itself  to
01900	record  observations and to collect data. This capacity of a computer
02000	allows us to develop a measurement standard , a basis  of  comparison
02100	in  which interaction patterns can be considered similar or different
02200	according to the measurement criteria.
02300		While   developing  a  computer-aided  treatment  method  for
02400	stimulating language behavior in nonspeaking autistic children (Colby
02500	and  Smith,1971,Colby,1973),  we became interested in the interaction
02600	patterns with the machine characteristic of these  children  and  how
02700	they  differed  from  normal children and others types of nonspeaking
02800	children. To evaluate the treatment method  we  attempted  to  follow
02900	interaction-pattern  changes  over  time  and to assess whether these
03000	changes could be considered as "improvement". If no changes  occurred
03100	or  if  the  changes  were  judged as "worsening", then the treatment
03200	should be discontinued. We chose to define improvement as a change in
03300	the  interaction pattern towards the pattern characteristic of normal
03400	speaking children in particular age groups.
03500		The   treatment   involves   letting  a  child  play  with  a
03600	computer-controlled   audio-visual    device    consisting    of    a
03700	typewriter-like  keyboard and a television-like video display screen.
03800	When a child presses a key, a symbol,  letter,  word,  expression  or
03900	drawing  appears  on  the screen accompanied by a sound, mainly human
04000	voice sounds and some animal or machine sounds. The  overall  program
04100	is divided into "games" of varying complexity. A sitter who sits with
04200	the child changes the games in accordance with the child's  interests
04300	and  abilities.  The simplest game is Game #1 in which pressing a key
04400	produces that key's symbol on the video screen accompanied by a voice
04500	pronouncing  the  corresponding  letter or number. We shall limit our
04600	discussion of objective measurement to the data collected in this one
04700	game.
04800		To keep track of the child's interactions with the machine, a
04900	program  was  written  (by  Earl Sacerdoti, a graduate student in the
05000	Department of Computer Science, Stanford University)  which  recorded
05100	the  game  the child was playing, which key was pressed at what exact
05200	time and whether the sound for this key was  played  over  the  audio
05300	device.  The  computer-controlled  system  is designed so   that if a
05400	child presses a single key or several keys in bursts of less  than  a
05500	second's duration, the sound for the first pressing is played but the
05600	rest are not in order to avoid confusing the child. As  soon  as  the
05700	child  pauses  at the end of such a burst, a buffer is cleared of all
05800	the symbols accumulated during the burst, and when the  next  key  is
05900	pressed, its sound is played. Striking the keys at an extremely rapid
06000	rate indicates a child is ignoring the  sound  and  paying  attention
06100	only to his visual and/or tactile experience.
06200		In Fig. 1 the first column on the left indicates  the  number
06300	of  the  game,  the second column indicates the key struck, the third
06400	column shows the time of striking in hours-minutes-seconds,  and  the
06500	fourth  shows whether the sound for that key was played ("effective")
06600	or not (blank space). The data of Fig. 1 illustrates the  interaction
06700	pattern  of  a  nine year old nonspeaking boy clinically diagnosed as
06800	autistic and with a score of +26 on Rimland's  E2  scale.   (A  score
06900	greater  than +20 is considered by Rimland to indicate a true case of
07000	early infantile autism (Rimland,1971).  Fig.  2  shows  data  from  a
07100	normal  17  month  old  speaking boy. It is noteworthy how, on simple
07200	visual inspection, the pattern of  the  9  year  old  autistic  child
07300	resembles  that  of  a  17  month  old  normal  child. In this way an
07400	objective measurement allows us  to  establish  equivalences  between
07500	different children.
07600		To analyze data from several children  playing  Game  #1,  we
07700	examined 6 numerical scores for each child - the total number of runs
07800	(a run consisting of the same key being pressed), the length  of  the
07900	run,  the  average  length of a run, the number of gaps (blank spaces
08000	indicating non-listening to sound), the length of the  gap,  and  the
08100	average  length of a gap. The data for these measures from normal and
08200	autistic children appear in Tables 1 and 2.
08300		From the normal children data we can  construct  a  curve  as
08400	shown  in  the graph of Fig. 3. Normal children from age 17 months to
08500	10 years progress in their interaction patterns along this curve. The
08600	interaction  patterns  of  a  nonspeaking  autistic  child locate him
08700	somewhere on this curve. Over time it can be estimated whether he  is
08800	progressing  towards  a more normal pattern, whether he has reached a
08900	plateau or whether he is retrogressing.   We have  not  yet  observed
09000	the  latter  case.  Since  the  data of the curve is longitudinal, it
09100	would  be  ideal  if  we  could  show  the  progress  in  normal  and
09200	nonspeaking  autistic groups of children over 10 year periods.  Since
09300	we have only cross-sectional data thus far, this is what we  used  to
09400	construct the curve.
09500		As yet we do not have complete data from start to  finish  of
09600	treatment  on  an  improved  case  of  a  nonspeaking autistic dhild.
09700	However, Fig.  4 shows the current (Aug.l973) interaction pattern  of
09800	J, a 10 year old boy whom we treated for two years and whose language
09900	development gained markedly.   (We were not collecting this  type  of
10000	data when he was in treatment.) Stretches of J's interaction patterns
10100	are quite normal looking and his location on the  interaction-pattern
10200	curve  can  be seen in the graph of Fig.3.      We have several cases
10300	of failure in which the final sessions of treatment show  interaction
10400	patterns  indistinguishable  from  those  of the early sessions.  The
10500	data of M. (see Fig. 1) reveal no change over the past two years.  We
10600	will  try  for  another  year  and  if  no  change  occurs,  we  will
10700	discontinue. We must also be prepared for the possibilities that  (1)
10800	a child's comprehension and speech improves as judged by clinical and
10900	parental observation but his interaction patterns do  not  change  or
11000	(2)  a  child's interaction patterns change towards the normal but he
11100	still does not use speech for social communication. As yet,  we  have
11200	not observed either of these paradoxical outcomes.
11400		In summary, we have presented  an  objective  measurement  of
11500	children's interaction patterns in playing with a computer-controlled
11600	audio-visual device programmed  to  stimulate  language  development.
11700	This  measurement  is  useful in 3 ways: (1) it reveals where a child
11800	stands on an interaction curve relative  to  normal  and  other  non-
11900	speaking  autistic  children; thus treatment can be planned suited to
12000	the child's position on the curve  (2)  changes  in  the  interaction
12100	pattern  over time can be evaluated to see if a child is improving or
12200	not, and (3) if no change takes place or a child  reaches  a  plateau
12300	and  remains  there for a long time, discontinuation of the treatment
12400	can be justified.   Thus  an  objective  measurement  of  interaction
12500	patterns  serves  as  a useful instrument in planning and terminating
12600	treatment of language deficiencies in nonspeaking autistic children.
12700	                          References
12800	
12900	Colby, K. M. The rationale for computer-based treatment of
13000	     language difficulties in nonspeaking autistic children.
13100	     Journal of Autism and Childhood Schizophrenia,l973, 3, 254-260.
13300	
13400	Colby, K. M. and Smith, D. C.  Computers in the treatment
13500	     of nonspeaking autistic children.  In J. H. Masserman
13600	     (Ed.), Current Psychiatric Therapies, Greene & Stratton,
13700	     N.Y., 1971.
13800	
13900	Rimland, B.  The differentiation of childhood psychoses: an
14000	     analysis of checklists for 2,218 children.  Journal of
14100	     Autism and Childhood Schizophrenia, 1971, 1, 175-189.