TPC-Journal-V2-Issue2

The Professional Counselor \Volume 2, Issue 2 110 Figure 4. Initially proposed model. e_sa e_ps Self Appraisal Problem-Solving Career Problem-Solving Self Efficacy e_cdse Career Exploratory Behavior Career Thinking Decision-Making Confusion Commitment Anxiety External Conflict e_ec e_ca e_dmc e_ct e_ceb Self- Exploration Environmental Exploration e_se e_ee 1 1 1 1 1 1 1 1 1 1 1 1 1 Measurement Model The measurement model was evaluated to determine how well the observed variables combined to represent the underlying latent constructs of the model (Anderson & Gerbing, 1988). The initial measurement model was constructed with three latent constructs: career exploratory behavior, career thinking and career problem-solving self-efficacy (See figure 4). For this step, all latent variables were allowed to co-vary, and observed indicators were restricted to load only on their respective factor. Career problem-solving self-efficacy’s observed variables, problem-solving and self-appraisal, were significantly correlated ( r =.71; p < .01). Career thinking’s observed variables—decision-making confusion (DMC), commitment anxiety (CA) and external conflict—were significantly intercorrelated ( r =.53-.71; p < .01). Exploratory behavior’s observed variables, self-exploration and environmental exploration, were significantly correlated ( r =.34; p < .01). The loadings of the measured variables on the latent variables were all statistically significant ( p < .001) and ranged from .55 to .95, providing evidence that each of the latent variables was adequately measured by their respective observed variables. Structural Models The initial model tested (See Figure 4) proposed a bi-directional relationship between the career thinking and career problem-solving self-efficacy latent variables, along with variables explaining direct variance in career exploratory behavior. During the analysis process in AMOS 16.0, the model was deemed inadmissible. Therefore, a second path