The Stepwise Hybrid Statistical Inference Strategy: FOSTIS.

Notwithstanding the body of venerable and formidabie criticism (e.g., Birnbaum, 7962; Bakan, 1966; Meehl, 1967; Morrison & Henkel, 1970), demand remains unwavering for inferential argument articulated via significance. To date there are several possible grounds for this persistence; perhaps foremost of these being a certain objective obtained from algorithmic qualification of results. Relieved of the burden of result qualification, research requires only the null hypothesis prediction. In this paper we stand with previous objection to the current standard by arguing that its usage has undesirable effects on theory development and therefore should be modified so that prediction takes a more specific form. Having discussed this and in view of other considerations, an alternative is put forward and discussed