Multinomial Regressions
Table 4 presents the multinomial logistic regression for occupational predictions. Compare to service occupations, people who are more extravert are more prone to work in education, education, legal, community service, arts, media, and military/ healthcare practitioners and technical/ farming, after controlling for other variables. Also, less conscientious people tend to work in construction and extraction/ production than service occupations. Only neuroticism has no effect on occupations. Similar patterns can also be seen in gender separated models (analyses not shown), but males with less conscientiousness are apt to work in sales/ office and administrative/ farming also. These findings support my hypothesis 2 that …show more content…
Net of the effects of control variables, the effects of non-cognitive traits are positively related to wages. But, the effect size and significance level of extraversion declines after controlling for other variables. And the effect sizes of neuroticism and conscientiousness both go up because of the suppression effect of high school GPA and blacks respectively. These results, again, support H1b and H1c, but not H1a. As for gender differences, the effect of neuroticism in females becomes significant after the inclusion of other variables, while the significant effect of extraversion in males diminishes after the control. These indicate that non-cognitive traits are more influential to females than to …show more content…
This indicates Model 16 fits better to the data. Also, other than non-cognitive traits, findings on cognitive traits and demographic variables also show some interesting patterns. First, females earn significantly lower than males do. This shows the ongoing gender inequality in the labor market. Secondly, for cognitive traits, both year of education and verbal IQ have positive impacts on wages. Thirdly, for races, though black and other races males tend to earn less, black women earn slightly higher than white women after controlling for other variables. Fourthly, males from single parent families and females from other family structures earn significantly lower than those from two parents families. Finally, the pattern of occupations on wages indicate that compare to service occupations, almost all other occupations earn higher wages, except for sales and related/ office and administrative supports/ farming, fishing, and
Table 4 presents the multinomial logistic regression for occupational predictions. Compare to service occupations, people who are more extravert are more prone to work in education, education, legal, community service, arts, media, and military/ healthcare practitioners and technical/ farming, after controlling for other variables. Also, less conscientious people tend to work in construction and extraction/ production than service occupations. Only neuroticism has no effect on occupations. Similar patterns can also be seen in gender separated models (analyses not shown), but males with less conscientiousness are apt to work in sales/ office and administrative/ farming also. These findings support my hypothesis 2 that …show more content…
Net of the effects of control variables, the effects of non-cognitive traits are positively related to wages. But, the effect size and significance level of extraversion declines after controlling for other variables. And the effect sizes of neuroticism and conscientiousness both go up because of the suppression effect of high school GPA and blacks respectively. These results, again, support H1b and H1c, but not H1a. As for gender differences, the effect of neuroticism in females becomes significant after the inclusion of other variables, while the significant effect of extraversion in males diminishes after the control. These indicate that non-cognitive traits are more influential to females than to …show more content…
This indicates Model 16 fits better to the data. Also, other than non-cognitive traits, findings on cognitive traits and demographic variables also show some interesting patterns. First, females earn significantly lower than males do. This shows the ongoing gender inequality in the labor market. Secondly, for cognitive traits, both year of education and verbal IQ have positive impacts on wages. Thirdly, for races, though black and other races males tend to earn less, black women earn slightly higher than white women after controlling for other variables. Fourthly, males from single parent families and females from other family structures earn significantly lower than those from two parents families. Finally, the pattern of occupations on wages indicate that compare to service occupations, almost all other occupations earn higher wages, except for sales and related/ office and administrative supports/ farming, fishing, and