Discussion

dence and classroom control are directly related to teachers’ levels of enthusiasm in their positions. Greater autonomy in a classroom, and more involvement in school-wide decisions, likely make teachers feel more involved and interested in their positions and stimulate a desire to try new interesting pedagogies and practices (Zuckerman et al., 1978). Such educator involvement can also have an effect on students, as research suggests that students who perceive their instructors as auton- omous experience more interest and enjoyment, as well as less anxiety, in classes (Black and Deci, 2000). Involvement in school decisions also directly influences teachers’ satisfaction and professional investment. The link between interdepen- dence and satisfaction is expected given research on procedural justice that suggests that involvement in a process (Thibaut and Walker, 1975, 1978) that represents parties’ interests (Leventhal et al., 1980) and promotes within-group relations (Lind and Tyler, 1988) increases satisfaction with both the procedure and outcome (Tyler et al., 1997). Similarly, previous research (Bolger and Somech, 2004) suggests that efficacy also increases commitment and organizational citizenship behaviors. The heavy influence of interaction on enthusiasm suggests that enthusiasm is more prone to external stimuli than satis- faction levels. It is reasonable to think that satisfaction in work may be influenced by an internal psychological state-of-mind, such as teacher self-efficacy (Bolger and Somech, 2004; Chan et al., 2008) whereas enthusiasm for one’s job is much more likely to be influenced by the surrounding ethos of the work environment. The uneven and opposite effects of enthusiasm on satisfaction and satisfaction on enthusiasm may indicate a cognitive dissonance effect (Festinger and Carlsmith, 1959) where people who report higher levels of satisfaction are psychologically compensating for lower than expected levels of enthusi- asm or, alternatively, it could be that these teachers who are satisfied are more likely to have resigned themselves to a less- enthusiastic and lower-pressure standpoint (Jensen, 2003). That interest increases satisfaction is unsurprising, as is the di- rect effect between interaction and cohesion (Friedkin, 2004). The final unexpected, yet logical, results were the direct effects of enthusiasm and satisfaction on teachers’ commitment to stay. It is reasonable that both interest in and satisfaction with a position makes one more likely to remain in that position net of any relational attachment predicted by the affect theory’s endogenous process (Van Dick et al., 2004). The effects of the control variables align with other research. Females tend to report higher levels job commitment overall (Abbott, 1993). The reasoning behind this finding is still unclear. Theories on the influence of relative perception (women comparing themselves to other women) and lower economic expectations have been purported, but not verified. Given the sample for this study, the ‘female’ effect may also be due to a homophily effect since elementary schools are heavily staffed by females. This homophily effect is also a likely explanation for the positive white, non-Hispanic effect on relational cohesion and affect processes. Homophily is a characteristic associated with stronger relationship formation and mainte- nance (McPherson et al., 2001). Moreover, the mostly negative proportion minority teachers in the school and urbanicity effects are additive since the vast majority of minority teachers teach in urban schools (Lippman et al., 1996). In sum, while the final model articulates a more complex relationship between power, commitment, and the intervening variables than the affect theory’s model, we see the extraneous effects as complementary rather than discouraging. The af- fect theory’s processes are certainly at work in school settings. Interdependence increases interaction, which raises positive affect that amplifies cohesion. This cohesion increases commitment. Although the process appears more robust with female teachers, the endogenous process appears to work for teachers across the spectrum (e.g., new and old, urban and rural, etc.).