If you know the size of the universe/population, you should opt for probability sampling wherein the sample size is determined scientifically. In such a scenario, you can opt for the Krejcie and Morgan formula and table (1970). You can refer this: https://www.kenpro.org/sample-size-determination-using-krejcie-and-morgan-table/#google_vignette
In retrospective studies, sample size calculation can be somewhat challenging but is not impossible; it is certainly more common to rely on convenience sampling due to the nature of the data being analyzed, which is typically collected for other purposes. However, if you have access to a sufficiently large and relevant dataset, performing a sample size calculation can enhance the validity of your findings. To calculate the sample size for a retrospective study, you would generally start by defining the primary outcome of interest and estimating the expected effect size based on prior research or pilot data. Next, you need to decide on the desired statistical power (commonly set at 80% or 90%) and significance level (typically 0.05), which will influence the required sample size. Additionally, you must consider the proportion of participants that might be lost to follow-up or the rate of incomplete data, as this will affect your final sample size. Using software or statistical formulas designed for sample size estimation can provide a more robust and reliable estimate, guiding you to select an adequate sample size that minimizes bias and allows for generalizability of your results. Ultimately, while convenience sampling may be practical, a well-thought-out sample size calculation can provide a stronger foundation for your analysis, especially if you can identify potential confounding factors and ensure that your sample is representative of the population of interest.