DIVIDED ATTENTION: PERFORMING TWO TASKS SIMULTANEOUSLY
-CAN WE DO BOTH TASKS AS EFFICIENTLY AS WE COULD DO EITHER TASK BY ITSELF

CONSIDER: SEARCHING FOR THE LETTER "K" IN A DISPLAY OF NUMBERS.

CONTRAST WITH SEARCHING FOR EITHER AN "X" OR A "Q"

OR: SEARCHING FOR EITHER AN "X" OR A "Q" OR A "B" OR AN "E" .....

IN THE EARLY DAYS THE ANSWER GIVEN TO THIS QUESTION WAS "NO" PERFORMING SEVERAL TASKS SIMULTANEOUSLY LEADS TO A COST (POORER PERFORMANCE ON AT LEAST ONE OF THE TASKS)

BUT THERE WERE INDICATIONS THAT SOMETIMES THIS WASN'T TRUE: FOR INSTANCE, ONE STUDY, SHOWED THAT A PERSON COULD BE TRAINED TO SEARCH FOR 10 LETTERS AS EFFICIENTLY AS SEARCHING FOR JUST ONE

IN THE MID-1970'S SHIFFRIN AND SCHNEIDER DID A CLASSIC EXPERIMENT THAT CLARIFIED CONDITIONS IN WHICH PEOPLE COULD DO TWO TASKS SIMULTANEOUSLY


SHIFFRIN AND SCHNEIDER ASKED PARTICIPANTS TO SEARCH FOR CERTAIN TARGETS (EITHER LETTERS OR DIGITS) IN DIFFERENT ARRAYS OF LETTERS OR DIGITS (THEY CALLED AN ARRAY THE FRAME. FOR EXAMPLE, A PERSON MIGHT BE ASKED TO SEARCH FOR THE LETTER "J" (THE MEMORY SET, A LA
STERNBERG) IN THE FOLLOWING FRAME B M K T. (NOTE; THIS IS BASICALLY THE SEARCH TASK USED BY TREISMAN THAT WE TALKED ABOUT EARLIER).

THERE WERE SEVERAL MANIPULATIONS; MAPPING FROM TRIAL TO TRIAL IN THE VARIED-MAPPING CONDITION THE ITEMS TN THE MEMORY SET COULD OVERlAP WITH ITEMS IN THE FRAME. THAT IS, TARGETS IN ONE TRIAL COULD BE DISTRACTORS ON SUBSEQUENT TRIALS.
e.g., TRIAL N SEARCH FOR J IN FRAME B M K T
-TRIAL N+ 1 SEARCH FOR M IN FRAME L J T Q

EXPECTED THAT THIS WOULD BE A DIFFICULT, ATTENTION DEMANDING TASK

IN THE CONSISTENT-MAPPING CONDITION THE MEMORY SET CONSISTED OF LETTERS AND THE FRAMES OF DIGITS, OR VIVE VERSA. THUS STIMULI THAT WERE TARGETS IN ONE TRIAL WERE NEVER ALSO DISTRACTORS IN OTHER TRIALS

EXPECTED THAT THIS WOULD BE AN EASY TASK, REQUIRING MUCH LESS ATTENTIONAL RESOURCES

OTHER MANWULATIONS:
FRAME SIZE: FROM TRIAL TO TRIAL VARIED FROM I TO 4 ITEMS
FRAME TIME: LENGTH OF TIME FRAME PRESENTED: 20 TO 800 MSEC
MEMORY SET SIZE: NUMBER OF TARGETS ONE IS LOOKING FOR (1-4)


THE BASIC FINDINGS:
WHEREAS, INITIALLY, PERFORMANCE IN BOTH THE VARIED AND CONSISTENT-MAPPING CONDITIONS EXHIBITED SIMILAR PATTERNS OF PERFORMANCE, AFTER EXTENSIVE PRACTICE ON DOING THE TASK, ONE FINDS CHANGES IN THE PATTERN FOR THE CONSISTENT (BUT NOT THE VARIED-MAPPING CONDITON)

AFTER 14 DAYS OF PRACTICE, THIS IS WHAT WAS OBSERVED.

FOR THE CONSISTENT-MAPPING CONDITION
-SUBJECTS PERFORMANCE VARIED WITH ONLY FRAME TIME.  THE LIKELIHOOD OF CORRECTLY SAYING A TARGET WAS PRESENT WHEN IT ACTUALLY WAS SPEED, DECREASED AS THE SPEED OF FRAME PRESENTATION WENT FROM 120 TO 40 MSEC.

-PARTICIPANTS WERE JUST AS ACCURATE SEARCHING FOR I AS FOR 4 TARGETS (MEMORY SET SIZE) OR WHETHER THERE WAS 1 OR 4 DISTRACTORS (FRAME SIZE). THAT IS, THE RT EXHIBITED A SLOPE CHARACTERISTIC OF PARALLEL PROCESSING

FOR THE VARIED-MAPPING CONDITION
-IN ADDITION TO THE EFFECT OF FRAME TIME, THERE WERE ALSO EFFECTS OF BOTH FRAME SIZE AND MEMORY SET SIZE

-THAT IS, THE LIKELIHOOD OF CORRECTLY SAYING A TARGET WAS PRESENT WHEN IT ACTUALLY WAS INCREASED IN A LINEAR FASHION AS MEMORY SET SIZE INCREASED AND, ADDITIVELY, AS THE NUMBER OF DISTACTORS INCREASED

-THAT IS, RT IN THESE CONDITIONS EXHIBITED THE INCREASING SLOPE OF A SERIAL SEARCH


BACK TO THE IDEA OF ATTENTIONAL RESOURCES:
IF WE HAVE ONLY A LIMITED POOL OF ATTENTION TO GIVE, HOW CAN THESE RESULTS BE EXPLAINED?

SHIFERIN AND SCHNEIDER ARGUED THAT THESE DIFFERENCES WERE DUE TO DIFFERENCES BETWEEN AUTOMATIC AND CONTROLLED PROCESSES

AUTOMATIC=> OPERATES IN PARALLEL (CAN BE DONE SIMULTANEOUSLY WITH OTHER PROCESSES) and DOES NOT CALL UPON ATTENTIONAL RESOURCES
-WOULD BE USED FOR EASY TASKS, AND FAMILIAR ITEMS
--->THAT IS, AS A FUNCTION OF PRACTICE

CONTROLLED ==> IS USED FOR DIFFICULT TASKS, AND THOSE INVOLVING UNFAMILIAR PROCESSES. IT OPERATES SERIALLY, REQUIRES ATTENTION, IS CAPACITY-LIMITED, AND IS UNDER CONSCIOUS CONTROL

SO WHEN SHOULD WE BE ABLE TO DO TWO TASKS SIMULTANEOUSLY?
IN THEORY, IF ONE TASK IS AUTOMATIC THEN ONE SHOULD BE ABLE TO DO ANY OTHER TASK AT THE SAME TIME WITHOUT COST (RELATIVE TO THE PERFORMANCE WHEN THE SECOND TASK IS DONE BY ITSELF)

HOWEVER AS SHOWN (SEE SPELKE ET AL STUDIES, P. 118) IT SEEMS THAT THE RELATIONSHIP OF THE TASKS BEING DONE SIMULTANEOUSLY IS QUITE IMPORTANT --- MUST BE ABLE TO COORDINATE THE VARIOUS TASKS, WHICH MIGHT INVOLVE DISCRIMINATING BETWEEN THEM AT SOME CRITICAL POINTS IN PROCESSING


Psych 235 Home Page