I have designed my questionnaire with 4 latent factors to measure peoples' perceptions about public spaces such as streets and squares. Each latent factor is being measured by about 5 indicators which i extracted from literature review.For example the latent factor of safety is measured by 1.presence of other users 2. presence of cameras 3.presence of police 4. low traffic speeds 5. Clearly marked crosswalks. Do I perform factor analysis for construct validity before performing further inferential statistics in analysis? Thanks