How does increasing the sample size ($n$) impact the standard error of a population mean, assuming the population standard deviation ($\sigma$) remains constant?