3.1.  Cost 1: Translating Type I errors into theory

3.2.  Cost 2: Failing Popper’s criterion of disconfirmability

3.3.  Cost 3: Disguising accommodation as prediction

3.4.  Cost 4: Not communicating information about what did not work

3.5.  Cost 5: Taking unjustified statistical license

3.6.  Cost 6: Presenting an inaccurate model of science

3.7.  Cost 7: Encouraging ‘fudging’ in other grey areas

3.8.  Cost 8: Making researchers less receptive to serendipitous findings

3.9.  Cost 9: Encouraging the adoption of narrow, context-bound theory

3.10.  Cost 10: Encouraging the retention of too-broad theory

3.11.  Cost 11: Inhibiting the identification of alternative hypotheses

3.12.  Cost 12: Violating the ethical principles of honesty and openness