Framework-Based Qualitative Analysis of Free Responses of Large Language Models: Algorithmic Fidelity