Grid questions become easier when you identify each marked cell by its row and column. A transformation can then be checked systematically instead of judged only by visual resemblance.
Fix the coordinate convention
For a 3 × 3 grid, number rows from top to bottom and columns from left to right. The top-right cell is therefore (row 1, column 3).
This row–column convention differs from a mathematical x–y plane with y increasing upward. State which convention you are using before applying a formula.
Rotate a marked cell
A 90-degree clockwise rotation of a 3 × 3 grid maps (row, column) to (column, 4 − row). A mark at (1, 2) moves to (2, 3).
Apply the same rule to every mark. If a shape contains several cells, transforming just its most noticeable cell is not enough to verify the whole answer.
Reflect across the vertical center line
A vertical reflection maps (row, column) to (row, 4 − column). Thus (1, 2) stays at (1, 2), while (2, 3) moves to (2, 1).
Marks on the mirror line remain in place. This is an easy check against an option that unnecessarily moves every cell.
Order can change the result
Start at (1, 2). Rotate clockwise, then reflect vertically: (1, 2) → (2, 3) → (2, 1).
Reverse the order: reflect first, then rotate. The point remains (1, 2) after reflection and then moves to (2, 3). The two results differ.
Preserve the whole pattern
After a rigid grid transformation, the number of marked cells and their relative distances should be preserved. Check those invariants along with the coordinates. For a translation, also verify whether moving beyond the boundary is prohibited or whether the puzzle explicitly allows wrapping; never assume wraparound without a rule.