In the past decade, there has been a noticeable increase in the number of courses/programs offered online by higher education institutions in US and worldwide. This increase makes it necessary to comprehensively study the effectiveness of these offerings to ensure that they yield comparable outcomes to traditional offerings. This research discusses findings of a multiyear, in-depth quantitative study with the objective of assessing the effectiveness of delivering computer programming courses online and analyzing the quality of this delivery mechanism. The paper addresses several aspects of comparing tradition/blended offerings with fully online offerings. Several data sets are used in this study with primarily focus on direct assessments. Specifically, this research measures the Intended Learning Outcomes (ILOs) attained by students enrolled in online programming classes and compares them to the ILOs obtained when teaching the same course in a blended format. Moreover, several other students' success and involvement indicators are measured in both cases and compared. Besides, the paper briefly addresses other indirect assessment findings. The employed data sets have been collected over a number of years (from 2006 till 2011) to enhance the accuracy and implications of the reported statistical findings. Generally, no significant differences were found regarding the attainment of almost all of the ILOs with the exception of one that showed better performance, on average, in the online offerings. On the other hand, blended offerings yielded favorable results that are statistically significant when considering other performance criteria such as withdrawal and failure rates.