Purpose: Various recent studies are perplexed with the situation if subjective measures such as perceptions of learning reflect knowledge gains and whether they can be used as a surrogate for evidence of actual learning in an online learning environment. Due to a rising trend in current research studies interpreting perceptions of learning compared to actual learning as a measure of success, this study investigates the dichotomous measures of objective and subjective measures of academic performance in an online learning environment with selected predicted variables. This is followed by an investigation on the effect of online learning on students' academic performance prior to and during the COVID environment. Methodology: This cross-sectional study employed a correlation design for the purpose of data collection among 382 university students. The dependent variable of study comprised objective measures of CGPA scores while the subjective measures was based on a composite score from perceived items in a questionnaire. The predictive independent variables included, (1) satisfaction towards online learning, (2) online learning self-efficacy, (3) social interactive engagement in online learning, and (4) online learning environment. Findings -- The findings showed an extremely weak and nonsignificant relationship between actual academic performance (GPA) and perceived academic performance in an online learning environment. A statistically significant increase was notable in students' academic performance through GPA scores before and during COVID-19 academic semester and this increase was prevalent among low and intermediate achievers. Apart from environment, all other predictors were significant predictors of perceived academic performance. Engagement was a predictor for the subjective measure of CGPA scores. Significance -- Subjective measures data should not be considered as knowledge gain, and it should be used cautiously as a surrogate for evidence of actual learning. Educational research studies should be designed using more objective measures of learning rather than subjective measures in isolation as dependent variable. This should be seen as a caveat by policymakers in decision-making to practise consciousness in terms of judgemental modes employed in measurement and to exercise caution in collecting, analysing, and interpreting subjective data.