Data Collectionhard
0:00.0
A nutrition researcher tests three diets (A, B, C) on 9 volunteers over 12 weeks, with each volunteer trying each diet for 4 weeks in different orders. The researcher suspects two confounding factors: individual volunteers naturally lose weight at different rates, and weight loss improves with time (period effect). A Latin square design is used, with each diet appearing exactly once per volunteer and once per time period. Why is this superior to a simple repeated-measures design without blocking?
A nutrition researcher tests three diets (A, B, C) on 9 volunteers over 12 weeks, with each volunteer trying each diet for 4 weeks in different orders. The researcher suspects two confounding factors: individual volunteers naturally lose weight at different rates, and weight loss improves with time (period effect). A Latin square design is used, with each diet appearing exactly once per volunteer and once per time period. Why is this superior to a simple repeated-measures design without blocking?