Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)




Six investigations ware conducted to determine the efficacy of employing modifications of the Multidimensional Evaluation Structure Analysis (MESA) methodology to the predictive task of assessing potential vocational outcome for mentally retarded adults and psychiatric patients. Five of the investigations dealt xvLth mentally retarded adults who were seeking or had obtained vocational placement positions. A sixth investigation dealt with psychiatric patients (inpatients and outpatients) and their probable community adjustment success.

Each of the six subject groups (referred to as stimulus-persons) were evaluated by their respective judge group from four different agencies. The SS-1 and SS-2 stimulus-person groups (N = 10 and 11, respectively) were institutionalized at a State School for the Mentally Retarded. The OTC-1 and OTC-2 stimulus-person groups (N = 15 and 11, respectively) were clients at an Opportunity Training Center facility designed to train mentally retarded persons in occupational and self- help skills. The SITS stimulus-person group (N - 8) were clients of a Sheltered Workshop designed to train mentally retarded persons in skills leading to competitive employment. The PU stimulus-person group (N = 14) was selected from a psychiatric unit functioning to assess and treat persons with psychiatric problems. The number of judges used for each investigation ranged from three to nine.

Predictor values were derived from factor loadings extracted by the MESA methodology. MESA is a "content" scaling methodology that requires a judge to estimate the perceived degree of similarity (using values from zero to 100) between two stimuli. In these investigations, the stimuli were persons (called stimulus-persons) in the groups described above. The judge groups x^ere asked to evaluate their respective stimulus-person group. Also, the judges were asked to indicate on a five-point scale how familiar they were xriLth each stimulus-person.

For the SS-2, 0TC-2, SWS, and PU groups, the judges were also asked to estimate the probable degree of success (using a value from zero to 100) that each stimulus-person might experience x

Txjo MESA computational routines were employed including the Group Composition Structure Analysis (GCSA) and the Observer Factor Judgment Analysis (OFJA). The GCSA found underlying judgmental criteria used by the judges where the OFJA found clusters of judges demonstrating a commonness of judgmental strategy. MESA predictor values were formulated from GCSA factor loadings and x-rere modified xtfith regard to how much variance each factor (judgment-dimension) accounted for and by how familiar each judge was xdLth the stimulus- persons. The predictor values were correlated with follow-up data. These relationships reflected the efficacy of using the MESA methodology in assessing clinical judgment and staff consensus as predictors of success outcomes for individuals in groups.

Three studies demonstrated that clinical judgment (using MESA and similarity estimations) can adequately predict future success with regard to the samples investigated. The three studies showing little or no predictive success had judges that used different judgmental strategies, as shown by the OFJA analysis.

The six investigations demonstrated that clinical judgment allows for adequate prediction provided the clinical judgments are handled in an objective and systematic way. Using MESA as a predictive format also allows the investigator to determine when prediction will probably fail based on the OFJA analysis. The study demonstrated a methodology that may be used to systematically assess staff consensus and individual clinical judgment in predicting outcome for a variety of groups.