I doubt that Fleming could have obtained a grant for the discovery of penicillin on that basis [a requirement for highly detailed research plans] because he could not have said, 'I propose to have an accident in a culture so that it will be spoiled by a mould falling on it, and I propose to recognize the possibility of extracting an antibiotic from this mould.
Random search for data on ... off-chance is hardly scientific. A questionnaire on 'Intellectual Immoralities' was circulated by a well-known institution. 'Intellectual Immorality No. 4' read: 'Generalizing beyond one's data'. [Wilder Dwight] Bancroft asked whether it would not be more correct to word question no. 4 'Not generalizing beyond one's data.