I'm not aware that Likert scoring or scales have ever been used in epidemiology. Likert scales are intended to quantify [personal] opinions of groups of people. However, they may be useful in what we refer to as syndromic [disease] surveillance, to get some "feeling" about whether a disease exists or not on any epidemiologic unit (i.e. an aquaculture operation) - an approach not dissimilar from "if it looks like a fish, smells like a fish, it's probably a fish" (forgive the pun). While syndromic surveillance is useful is scanning large populations (global or countries) for emerging diseases, a simpler, more useful and direct approach on an aquaculture operation is, if there are signs of any disease, sample the the population and get get some diagnostics done quickly to confirm or reject the suspicion.
Thanks David Scarfe. You are right about the use of diagnostics. However, this method is applicable if sampling is being done presently. However, I am looking at history of disease occurrence where farmers have no records but may have an idea or perception of frequency of occurrence and measures they took that may include a call for diagnosis.
My idea is that you can use the Likert scale to quantify the information gain from the respondent. I guess that you need to quantify the past situations and therefores, this can be used to quantify the information/ data distilled from the respondents (fish farmers or other people who knows the history).