| A cross-sectional aging study was performed to examine the role of processing resources believed to underlie spatial visualization performance and its age-related decline. Measures of spatial visualization, working memory, general control processes, reaction time (RT) speed, attention, verbal ability, and verbal-analytic ability were administered to younger and older adult age groups, with the goal of understanding how these processing resource mechanisms explain tasks of spatial visualization, and their apparent age-related decline. Two cognitive interventions were administered during a spatial visualization test of rotation and folding in order to understand the modifiability of performance.; Psychometric modeling techniques were used to decompose the spatial visualization task into working memory and general control process components, and to estimate performance modifiabilities. Structural equation modeling techniques simultaneously estimated the influence of several cognitive resources on spatial performance.; The results indicated significant correlations two modifiabilities. Three series of models were fit to the data. The first series structural model results showed that the RT Speed factor accounted for spatial performance, but in the context of the final model, RT Speed had no mediating or direct effect on spatial visualization ability. In addition, age significantly Predicted Spatial Visualization performance after Attention, RT Speed, and Working Memory factors were estimated. The second series structural model results showed that psychometricaily-modelled components of Working Memory capacity and General Control Processes accounted for spatial performance over all other factors. Age had no additional prediction in Spatial Visualization after the other processing factors were estimated. The third series structural model results indicated that Verbal and Verbal-analytic ability mediated most modifiability and training performance variables.; Overall, the results support the use of psychometric item response models for task decomposition and modifiability estimation in understanding cognitive aging phenomena. |
