Experiments and studies commonly use random samples because observing an entire population is often too difficult, costly, time-consuming, or impractical. In most real-world settings, researchers cannot collect data from every individual, item, or event of interest. A properly selected random sample allows them to estimate population characteristics with a manageable amount of effort while still preserving the ability to make statistical inferences. Random sampling improves representativeness and reduces selection bias, but it does not guarantee total elimination of bias. That is why the option claiming bias elimination is incorrect. Likewise, the issue is not mainly about outliers or software cost. The main reason for using random samples is feasibility combined with inferential validity. With sound sampling methods, researchers can use probability theory to generalize findings from the sample to the broader population and estimate the degree of uncertainty in those conclusions. Therefore, the strongest and most accurate answer is that studying the entire population is difficult and impractical, making random sampling the preferred and efficient alternative.