Problem-solving by traumatically brain injured and neurologically intact subjects on an adaptation of the twenty questions test

Primary objective: RAPS (Rapid Assessment of Problem-Solving) is a clinical measure for assessing verbal problem-solving in hard-to-test patients or those that may not be able to tolerate a longer, more detailed assessment. The design of the test is based on Mosher and Hornsby's Twenty Question test, but RAPS contains several modifications to facilitate its use with brain-injured individuals. This study used RAPS to compare the verbal problem-solving ability of subjects that were neurologically intact and subjects that had chronic traumatic brain injuries. Methods and procedures: Twenty-one adults that were neurologically intact (NI) and 21 adults that had incurred a traumatic brain injury (TBI) matched for age, gender and education took part in the study. Before being tested with RAPS, participants signed an IRB-approved consent form and completed a battery of neurocognitive measures. RAPS entailed the solving of three verbal problems. Each problem involved an array of 32 pictures of common objects (e.g. football) arranged in a 4 × 8 grid. The subjects were instructed to ask yes/no questions to determine which picture the examiner was 'thinking of'. Three scores were computed for each problem solved: number of questions asked, percentage of constraint-seeking questions, and question-asking efficiency scores for the first four questions. Outcomes: No learning effects across the problems were found for any of the RAPS measures. Scores were averaged across the three problems to determine group effects. Groups of TBI and NI subjects did not differ significantly in the number of questions asked in solving RAPS problems. Members of the NI group asked significantly more constraint-seeking questions (e.g. Is it an animal?) than those in the TBI group, and the subjects that had incurred brain injuries did more guessing than the NI group. Over 70% of the time, guessing took place after the semantic category containing the target picture was known to the subject. Guesses took the form of pseudo-constraint questions (e.g. Is it the animal with a long neck?) rather than frank guesses (e.g. Is it the giraffe?). These trends were seen for both groups. Question-asking efficiency scores, computed for the first four questions of each problem, reflected the amount of information gained by the subjects' questions. It was anticipated that subjects' questioning strategies would target larger rather than smaller number of pictures and systematically reduce the number of total pictures under consideration. Question-asking efficiency scores were significantly higher for the group of NI subjects. Both groups increased question-asking efficiency scores across the first four questions, and there was no significant group × question interaction. Further analysis of the question-asking efficiency scores revealed that questions from the group of NI subjects tended to target multiple categories of pictures and larger single semantic categories of pictures on the 32-item problem-solving board, whereas those from the group of TBI subjects often targeted smaller categories or portions of categories. Conclusions: Two meta-cognitive functions, planning and strategy shifting, appeared to explain most of the differences in the verbal problem-solving performance between the groups. Both groups, however, demonstrated a range of abilities on RAPS. Until a larger normative database for RAPS is available, it behooves clinicians using the test to analyse results on an individual basis, to consider the subject's pre-morbid problem-solving ability and to weigh those factors associated with brain injury that could affect RAPS performance.