Correctness-by-construction (CbC) is a refinement-based methodology to incrementally create formally correct programs. Programs are constructed using refinement rules which guarantee that the resulting implementation is correct with respect to a pre-/postcondition specification. In contrast, with post-hoc verification (PhV) a specification and a program are created, and afterwards verified that the program satisfies the specification. In the literature, both methods are discussed with specific advantages and disadvantages. By letting participants construct and verify programs using CbC and PhV in a controlled experiment, we analyzed the claims in the literature. We evaluated defects in intermediate code snapshots and discovered a trial-and-error construction process to alter code and specification. The participants appreciated the good feedback of CbC and state that CbC is better than PhV in helping to find defects. Nevertheless, some defects in the constructed programs with CbC indicate that the participants need more time to adapt the CbC process.