Certain psychosocial characteristics of adolescents, such as self-esteem, perceived academic concept, educational expectations, and aggression are significantly related to high school completion or high school dropout. Finn and Rock (1997) found that students with higher self-esteem distinguished resilient students who drop out of school in comparison to those who remain in school in spite of poor grades and/or test scores; this finding is consistent across population race and gender subgroups, and when SES and family structure were controlled for. Leventhal, Graber, and Brooks-Gunn (2001) found that Black urban adolescents with higher levels of perceived academic ability at ages 16 to17 were more likely to graduate from high school than adolescents with lower levels of perceived academic ability. Along with other psychological indicators, educational aspirations have been found to predict educational attainment. Using longitudinal data on a mostly White suburban sample of students from upstate New York counties who were aged 5 to 10 when the study began and aged 21 to 26 at the last of four follow-ups, French and Conrad (2001) found that and earlier educational aspirations, at ages 15 to 20, predicted later academic orientation, as indicated by educational aspirations, GPA, and educational attainment. These results held after controlling for gender and academic achievement. Most of the sample was White, so race did not need to be controlled for. However, the SES of the sample was diverse and SES did not appear to be controlled for. Adolescents with higher levels of school attendance, who expected to attend college, and who reported to have positive attitudes toward school were less likely to have dropped out of high school and more likely to have received a diploma or GED than students with lower levels of attendance and less positive attitudes toward school (Astone and McLanahan, 1991). Another study found that students who were frequently absent when they were not ill had higher levels of high school dropout than those who were absent less frequently (Marsh, 1991b).

Psychosocial factors have been found to predict likelihood to attend college among adolescents. Marsh (1991), in an analysis of High School and Beyond, found psychosocial and factors to be independently predictive of higher levels of college attendance including educational aspirations and academic self-concept. Zaff, et al. (2001) found that adolescents' locus of control predicted their likelihood of attending college. These finding held after controls were added. Likewise, Borus and Carpenter (1984) found that high school seniors who reported that they expected to attend college were much more likely to attend college than those with lower educational expectations, after controlling for background differences. In a model with only pre-college and family level variables, Adelman (1999) found that educational aspirations were associated with higher likelihood of college completion. Yet, after college performance and attendance patterns were taken into account after students began attending college, prior educational aspirations were not significantly predictive of college completion.