Mutilaciones de ganado
19 REBIRTHING: EL OJO IZQUIERDO DE HORUS
The dependent variables of the present study were the scores of modified western neuropsychological tests. These tests were assumed to be more culturally sensitive than their original form because verbal and written english language instructions were replaced by a verbal aboriginal english dialect. Moreover, tests were embedded in the context of a story to elicit appropriate western tactics to solve the problems posed by each test. Most of the tests were modified in this way by using an animated format with interactive computer graphics and automatic data collection features. Apple Macintosh computer hardware (ie. SE/30) and software (ie. HyperCard) were used to develop a series of graphics cards for the basic test material. The resulting computer tests were administered to patients on a less expensive Apple Macintosh Plus computer with a 2MB RAM memory extension. Patients interacted with the com puter by spatially manipulating a com puter mouse. D etails of the software development and implementation have been outlined in Cocks and Berrie (1992). NP testing usually began with the mouse fam iliarisation task, followed by the aboriginal maze, card match, speed test and finally the memory screening test. This sequence was som etim es broken by necessary hospital routines but was usually resumed within 24 hours.
Familiarisation Task
The co m p u ter fam iliarisatio n task involved an anim ated graphics card of a scene w hich contained the elem ents of the m aze test (ie. stick figure of a child and rocks in a pool of w ater) and other fam iliar objects (ie. house, tree, h o rse and dog). P a tien ts w ere tra in e d to m ove the c u rso r to d ifferen t p o sitio n s on the com puter screen by encouraging them to m ove the co m p u ter m ouse. W hen the m ouse sw itch w as clicked, a sm all stick figure of a ch ild w ould au to m atically m ove to the last p o sitio n of the cursor. By using the m ouse in this way, p atien ts w ere asked to place the child on the horse, near the house, up the tree or on the spot of the dog's back. P atien ts w ere considered to have m astered the task if they were able to p o sitio n the child correctly on a rock in the pool o f w ater. They were then in tro d u ced to an anim ated m aze train in g card on w hich they had to help a sm all child cross a narrow creek along an invisible pathw ay in an array of 9 rocks. From this exercise patients learnt the rules of the m aze test and becam e fam iliar w ith the resu lts o f their in teractio n w ith the c o m p u ter. A fter rela tiv e ly little p rac tic e on the train in g card m ost p atien ts w ere able to in d ep en d en tly co n tro l the m ouse and com plete the tests. T w o o ld er men had problem s w ith m anual d ex terity w hich caused them to take too long to respond with the m ouse. To save tim e they were asked to po in t to their selected m oves w ith a thin stick w hile the w riter m an ip u lated the m ouse. This resu lte d in one of them "hunting" for the solution from the tester during his first m aze trial (M cE lw ain & K earney 1970, p35). P o o r perform ance on the fam ilia risatio n ex ercise resu lted in an a ssessm en t of the subject's visu al acu ity using a sm all p ictu re card. The p a tien t w as excluded from the study if he failed to m atch paintings on the back o f a truck with those on w indow s o f a house. The paintings and w in d o w s c o rresp o n d ed to the p attern s fro m the C ard M atch T est. O nly one p a tie n t fa ile d this test and he w as c o n firm e d b lin d by a p h y s ic ia n .
Aboriginal Maze Test
This was a computerised serial maze learning task designed to assess planning ability to code and retrieve visual spatial memory. The test represented further development of the hard wired Austin push button maze with its Milner pathway and resembled Barker's original stepping stone maze (Lezak 1983; Milner 1965, cited in Bowden 1988b; Barker 1931, cited in Walsh 1985). Individuals with ARBD generally use inefficient strategies to solve this learning task and although they eventually succeed in tracing the invisible path of the maze, they make more errors than normal subjects by repeating mistakes (perseverative errors) and failing to consistently follow the rules of the test (Lezak 1983; Walsh 1987). Even though individuals with ARBD learn to improve their performance with practice, they fail to m aintain error free trials well after the 10-15 trials normally required to reach an error free criterion (Walsh 1985). Damage to the frontal cortex in the form of atrophy or lesions are the usual pathological explanations for this performance deficit (Walsh 1985, 1987). The aboriginal maze test was presented on a computer graphics card as a flooded creek with an array of 100 rocks exposed above the water. After learning the rules of the test from the training exercise, the patient was required to use the computer mouse to take a group of 10 stick children one at a time across the creek to their camp on the other side. Most of the rocks were loose and would cause any child stepping on them to fall into the water. Other rocks represented the invisible Milner pathway between the banks of the creek. The patient was urged to learn the correct path as soon as possible to avoid all but the first child falling in the w ater (Bow den 1990, personal com m unication). The rem aining children w ere described as poor swimmers who could drown if they fell into the flooded creek. The test was adapted from a computer version containing written and auditory feedback about errors (Sutton 1990, personal communication; Cocks & Berrie 1992). Verbal instructions were consistent with the animated story but were similar in concept to those outlined in Bowden (1988a,b). As
with the maze training exercise, cues for a successful move were shown by the appearance of a child on the chosen rock. An unsuccessful move resulted in an auditory beep and the child appearing upside down in the water where the loose rock had been. After about 1 second the child was autom atically returned to the last safe rock. This visual m odification was c o n sid ere d to be e q u iv a le n t to the fin g er p o s itio n cues available to subjects operating the original bolt version of the stepping stone maze and the switch board of the Austin maze. Once a child had reached the last rock on the path, he autom atically w alked inside a house and the r e m a in in g c h ild r e n r a i s e d th e ir arm s as an a n im a te d g e stu re of a p p re cia tio n for help in g him across the cre ek . The p ro ced u re was repeated for each child over 10 separate trials. N um erical feedback about the num ber of errors per trial was not usually given unless patients had c o m p le te d the test and r e q u e s te d the in fo rm a tio n . S p ecial o p erating features were incorporated in the test to give more scoring options and if n e c e ssa ry en ab le e arly te rm in a tio n o f the test. S c o rin g was fully autom ated and included num ber of errors per trial, tim e taken per trial and n u m b er of errors around two vuln erab le tu rn in g points where the p ath tu rn e d aw ay fro m the goal. T h e s e m e a s u r e s p ro v id e d the d e p en d e n t v a ria b le s for the m aze test and in c lu d e d total maze error (TME), total maze perform ance speed (TMS) and total maze position error (M PE). Errors and speed of perfo rm an ce on the first trial were also included to help differentiate maze perform ance (ie. TR 1M E and T R IM S). Patients who appeared to be stuck at the vuln erab le turning points (ie. c re a tin g M P E) w ere g iven s u ffic ie n t p ro m p tin g to e n a b le th em to ev en tu ally c o n tin u e along the path. H ow ever, in c o n tra st to B o w d en ’s instru ctio n s, these prom pts were given at the v u ln era b le turning points rather than in unrelated sections of the maze. D ata conversion software was d e v e lo p e d to m a tc h c o m m e r c ia lly a v a ila b le s o ftw a re for the g rap h ic al and sta tis tic a l p re s e n ta tio n of re s u lts . D e sp ite the usual practice of m easuring the num ber of trials to c riterio n , the duration of testing was ratio n alize d in this study by allow ing p atien ts to com plete