Strategies vs Backward Induction in Dynamic Decision-making: An Experimental Investigation