Selection sort could be more efficient than bubble sort for large arrays of big elements due to the number of swaps performed during the sorting process.
In bubble sort, a swap is performed every time a comparison determines that two elements are in the wrong order. Since a swap of big elements (which contain a lot of data) can be costly in terms of memory operations, bubble sort can become less efficient because it potentially makes many such swaps throughout its sorting process.
On the other hand, selection sort performs the minimum number of swaps: it looks for the smallest (or largest, depending on the sorting order) element in each pass and makes exactly one swap at the end of each pass — it swaps the smallest found element with the element at the current position. So, even though selection sort may perform the same number of comparisons as bubble sort, it minimizes the number of swaps. When the elements are large and each swap is expensive, the reduced number of swaps can lead to better performance in terms of time complexity for selection sort compared to bubble sort.