C-179. Type 3 and Type 4 Errors
Efforts to find orderings already or previously there (C-93: Kf) are susceptible to three kinds of error, the first and second pertaining to inferences of nonchance relationship: erroneously accepting or not accepting the null hypothesis that no order has been observed. (The null hypothesis restates the envisioned ordering of, say, two conditions such that the research works as a screening tactic, to dismiss orderings that cannot even beat chance – let alone come close to affording a firm Grasp of whatever relationship obtains or might obtain.)
Type 3 error in Kf research occurs when the supposed relationship was the wrong focus of attention. The hypothesis was wrong to start with, whether it was subjected to the criterion of nonchance or to the “causal” criterion of, say, necessary and sufficient condition. The latter criterion has very limited application re the (Kt) orderings of human agency. Thus correlation is widely used there criterially—this despite the behavioral demands of problem solving – i.e., Realization (App. XIX; C-111) in accord with the needed functionality (C-144) imposed by the Nature of Things (III). Kt says we find out by trying.
“Trial and error” Tells us that Kt has been and still is no easy matter … that proposed relationships (aka hypotheses, theories, models) are riddled with type 3 errors. Trying is not easy. Nor is it fairly evaluated (C-16). (Thus the suggestion in App. IX that functional efforts be subjected to analysis as accidents rather than as failures.)
Type 4 error is to not try (VI). Whether by limiting oneself to conducting only Kf research, whether by limiting oneself to learning to the exclusion of knowing (XI; C-148: L/K>4+), whether by neglecting Realization (i.e., thinking that freedom TO makes no demands on capability development), whether just not trying … period.
(c) 2017 R. F. Carter
Type 3 error in Kf research occurs when the supposed relationship was the wrong focus of attention. The hypothesis was wrong to start with, whether it was subjected to the criterion of nonchance or to the “causal” criterion of, say, necessary and sufficient condition. The latter criterion has very limited application re the (Kt) orderings of human agency. Thus correlation is widely used there criterially—this despite the behavioral demands of problem solving – i.e., Realization (App. XIX; C-111) in accord with the needed functionality (C-144) imposed by the Nature of Things (III). Kt says we find out by trying.
“Trial and error” Tells us that Kt has been and still is no easy matter … that proposed relationships (aka hypotheses, theories, models) are riddled with type 3 errors. Trying is not easy. Nor is it fairly evaluated (C-16). (Thus the suggestion in App. IX that functional efforts be subjected to analysis as accidents rather than as failures.)
Type 4 error is to not try (VI). Whether by limiting oneself to conducting only Kf research, whether by limiting oneself to learning to the exclusion of knowing (XI; C-148: L/K>4+), whether by neglecting Realization (i.e., thinking that freedom TO makes no demands on capability development), whether just not trying … period.
(c) 2017 R. F. Carter
S