DESIGN RESEARCH FOR SAMPLE SIZE & POWER SOFTWARE

THE RESEARCH GOAL IS TO DEVELOP A PC WINDOWS BASED SOFTWARE PRODUCT TO FACILITATE THE PLANNING OF SAMPLE SIZE AND POWER BY RESEARCHERS THEMSELVES AS WELL AS BY STATISTICIANS. APPROPRIATE SAMPLE SIZES FOR HUMAN AND ANIMAL STUDIES ARE NECESSARY TO SUPPORT THE VALIDITY OF RESEARCH CONCLUSIONS AND TO ASSURE THE EFFICIENT USE OF RESOURCES. INVESTIGATORSMAY BE IGNORANT OF HOW TO SELECT A SAMPLE SIZE AND STATISTICIANS FACE A LARGE SCATTERED LITERATURE, INCOMPLETE TABLES AND SOFTWARE THAT IS HARD TO USE. THE PHASE I PLAN IS TO DESIGN A MODULAR INTERACTIVE SOFTWARE SYSTEM WHICH COMBINES A STRUCTURED INQUIRY AND RESPONSE FORMAT, TO ASSISTINVESTIGATORS IN SPECIFICATION OF STUDY DESIGN, OUTCOME VARIABLE TYPE, AND ANALYSIS GOALS, WITH COMPUTATIONAL AND GRAPHIC AIDS TO AID SPECIFICATION OF EFFECT SIZES AND VARIABILITY ESTIMATES. PRINTABLE TABULAR ENTRY/DISPLAY SCREENS AND INTERACTIVE GRAPHICAL DISPLAYS WILL ALLOW EASY ASSESSMENT OF MULTIPLE SCENARIOS. ALGORITHMS WILL BE CHOSENBASED ON COMPARATIVE EVALUATIONS. HELP SCREENS WILL PROVIDEDOCUMENTATION OF FORMULAS AND REFERENCES, AND INTERACTIVE DESIGN TUTORIALS. IN PHASE I, SPECIFICATIONS FOR MODULE 1 COMPONENTS FOR COMPARISON OF TWO PROPORTIONS WILL BE DETAILED, A DEMONSTRATION MOCK-UP OF THESE COMPONENTS PROGRAMMED, AND FEEDBACK OBTAINED FROM POTENTIAL USERS TO MODIFY AND IMPROVE THE SOFTWARE DESIGN. THE RESEARCH GOAL IS TO DEVELOP A PC WINDOWS BASED SOFTWARE PRODUCT TO FACILITATE THE PLANNING OF SAMPLE SIZE AND POWER BY RESEARCHERS THEMSELVES AS WELL AS BY STATISTICIANS. APPROPRIATE SAMPLE SIZES FOR HUMAN AND ANIMAL STUDIES ARE NECESSARY TO SUPPORT THE VALIDITY OF RESEARCH CONCLUSIONS AND TO ASSURE THE EFFICIENT USE OF RESOURCES. INVESTIGATORSMAY BE IGNORANT OF HOW TO SELECT A SAMPLE SIZE AND STATISTICIANS FACE A LARGE SCATTERED LITERATURE, INCOMPLETE TABLES AND SOFTWARE THAT IS HARD TO USE. THE PHASE I PLAN IS TO DESIGN A MODULAR INTERACTIVE SOFTWARE SYSTEM WHICH COMBINES A STRUCTURED INQUIRY AND RESPONSE FORMAT, TO ASSISTINVESTIGATORS IN SPECIFICATION OF STUDY DESIGN, OUTCOME VARIABLE TYPE, AND ANALYSIS GOALS, WITH COMPUTATIONAL AND GRAPHIC AIDS TO AID SPECIFICATION OF EFFECT SIZES AND VARIABILITY ESTIMATES. PRINTABLE TABULAR ENTRY/DISPLAY SCREENS AND INTERACTIVE GRAPHICAL DISPLAYS WILL ALLOW EASY ASSESSMENT OF MULTIPLE SCENARIOS. ALGORITHMS WILL BE CHOSENBASED ON COMPARATIVE EVALUATIONS. HELP SCREENS WILL PROVIDEDOCUMENTATION OF FORMULAS AND REFERENCES, AND INTERACTIVE DESIGN TUTORIALS. IN PHASE I, SPECIFICATIONS FOR MODULE 1 COMPONENTS FOR COMPARISON OF TWO PROPORTIONS WILL BE DETAILED, A DEMONSTRATION MOCK-UP OF THESE COMPONENTS PROGRAMMED, AND FEEDBACK OBTAINED FROM POTENTIAL USERS TO MODIFY AND IMPROVE THE SOFTWARE DESIGN.