Sample Size Required for Various Methods of Assessing Bone Status in Commercial Leghorn Hens

Abstract

A study was conducted to determine the appropriate sample size required for various methods used to assess tibial bone status in commercial Leghorn hens. The methods used were in vivo bone mineral content (BMC), in vivo bone density (BD), in vitro BMC, in vitro BD, tibia bone breaking strength (TBS), and percentage bone ash (BA). Dietary total P levels of .4, .45, .5, .55, and .7% were used as treatment source of variation. Twenty hens were sampled randomly to represent each dietary treatment. The CV for each bone status comparison method was estimated and was used in a procedure to estimate the sample size requirement for detecting a difference of δ between treatments.The sample size required to detect the difference between treatment means varied depending on 1) the method used to compare bone status; 2) the difference between the treatment means to be detected as significant (δ); and 3) the level of significance (α) assumed. The sample size required for various methods are tabulated at .01, .05, and .1 level of significance and for 2.5, 5, 7.5, 10, 15, and 20% δ. To detect an actual difference of 5% from the mean to be significant, at the .05 level of significance, a sample size of 44, 22, 31, 23, 47, and 85 hens per treatment would be necessary for in vivo BMC, in vivo BD, in vitro BMC, in vitro BD, TBS, and BA methods, respectively. The estimated sample size values would help researchers in designing experiments that involve bone status comparison of commercial Leghorn hens.