<![CDATA[This study aims to determine the effects of compensation and absenteeism on work productivity at PT Rumah Ice Cream Indonesia in South Jakarta, both partially and simultaneously. This study uses an associative quantitative method. The sampling technique uses a saturated sample of 62 respondents. Data analysis techniques used include data instrument testing: classical assumption testing, simple linear regression, multiple linear regression, correlation coefficient, and coefficient of determination. Hypothesis testing uses t-tests and F-tests. The results of the analysis show that compensation partially affects work productivity with a simple linear regression equation of Y = 23.016 + 0.383 X1. The correlation coefficient value of 0.663 indicates a strong relationship, and the coefficient of determination of 0.439 indicates that work productivity is influenced by compensation by 43.9%. This is proven by the t-test, which obtained a t-value > t- table (5.569>1.670) and reinforced by a significance value < 0.05 (0.000 < 0.05). The level of absenteeism partially has a positive effect on productivity, as shown by the simple linear regression equation Y = 23.016 + 0.214 X2. The correlation coefficient value of 0.520 means that there is a moderate relationship, and the coefficient of determination of 27% means that work productivity is influenced by the level of absenteeism by 27%. This is proven by the t-test, which obtained a t- value > t-table (3.192>1.670) and is reinforced by a significance value < 0.05 (0.002 < 0.05). The results of the study on compensation and absenteeism simultaneously influential affecting employee performance are shown by the multiple linear regression equation Y = 23.016 + 0.383X1 + 0.214X2. with a correlation coefficient of 0.722, meaning that there is a very strong relationship, and a coefficient of determination of 0.522, meaning that work productivity is influenced by compensation and absenteeism by 52.2%. This is proven by the results of the F test hypothesis, which obtained a calculated F value > F table (32.154> 3.156) and reinforced by a significance value < 0.05 (0.000 < 0.05). Thus, it is proven that Ho is rejected and Ha is accepted.]]>
