Randomness in Survey Responses

Imagine you're analyzing survey data composed of multiple-choice questions. How would you determine if certain individuals answered the survey randomly rather than providing honest responses?