Regression of attitude toward quitting drinking during pregnancy on factor scores showed that when women believed that quitting drinking would result in better health outcomes for their baby and themselves they had a more positive attitude toward quitting (β = .39, p <.0005). When women believed that quitting would increase their stress they had a negative attitude toward quitting (β = -.24, p =.001). See Appendix A for the path model showing the Pearson correlation and factor scores for attitudes toward abstaining. Predictors of Subjective Norm and Perceived Control
Factor analysis was unsuccessful in identifying a simple structure for the normative belief and perceived control variables probably due to a limited sample size and high multicollinearity. The simple Pearson correlation for normative beliefs and control beliefs did provide information relevant to one’s intention to quit. Appendix A shows a path model illustrating the association of the outcome, normative, and control belief optimally scaled scores with their corresponding theory of planned behavior variables. The path model shows that the five normative referents were highly predictive of the subjective norm to quit. All items for control beliefs were statistically significant predictors for perceived control to quit.

Differences between women who continued to drink and those who abstained when examining intention to quit drinking alcohol during pregnancy, based on the theory of planned behavior variables, significant differences between those who quit drinking during pregnancy and those who continued to drink were identified.

Women who continued to drink during pregnancy exhibit drinking patterns associated with problem drinking and/or addiction as shown in Table 3. It was also found that women who quit were closer to the end of their pregnancy. The following percentages represent the expectant due dates comparing those who quit vs. those who continued to drink, respectively: a) three or less months from now (76% vs. 24%), b) four to six months (42% vs. 58%), and c) more than six months (33% vs. 67%).

Summary of Result Findings
All three variables of the theory of planned behavior predicted intention to quit drinking alcohol. The correlation for attitude was .61 (p = .000) followed closely by perceived control .60 (p = .000). The subjective norm also had significant correlation with the intention to quit drinking alcohol during pregnancy .56 (p = .000). Once the correlation values were corrected for attenuation the values for each theory variable were .77, .80, and .89 for subjective norm, attitude, and perceived control respectively, indicating that perceived control best predicts intention. Calculations of attenuation corrected correlations used the Cronbach Alpha’s values of each scale as estimates of reliability and used the standard formula as found in StatSoft, Inc. (2006) among other places. Attitude was best predicted by the underlying factor of beliefs associated with improved health .39 (p = .000). All normative beliefs predicted the subjective norm well. Resources to reduce stress and increase one’s self efficacy in quitting best predict perceived control. The R2 for the multiple regression, indicating how effectively the theory of planned behavior variables predict intention to quit was .55 (p = .000). In addition, significant differences were found between the drinking patterns of women who quit and those who continued to drink.