Question

Suppose a study is being planned that w investigate whether female Beagles with severe periodontitis (gum and mouth disease) give birth to smaller litters of puppies, on average, than Beagles without periodontitis. Based on previous research, the standard deviation of litter size is estimated to be 2.4 puppies. Suppose we suspect that Beagles with periodontitis will give birth to on average 6 puppies whereas Beagles without periodontitis will give birth to on average 5 puppies a. Using the assumed values above, calculate Cohen's d. If you get a negative value, report it as positive. d= .417 b. This value (Cohen's d) can be thought of as... (select all that apply) The effect size The probability we will fail to reject the null hypothesis The assumed difference between mean litter sizes for Beagles with and without periodontitis, in terms of number of standard deviations The probability we will reject the null hypothesis The standard error of the effect size The assumed difference in standard deviations of litter size for Beagles with and without periodontitis, in terms of the means c. Let's assume that population effect size is Cohen's d 0.5 (in practice the population effect size is unknown, so we just have to assume a value). For each value of desired power, find the total sample size (i.e. the sample size for both groups combined) required to detect this effect. Use the "Sample Size and Power" tool in JMP, found under the DOE/ Design Diagnostics menu. If you have trouble, take a look at the class example on power. i. For power 0.55, required n254 ii. For power 0.8, required n i. For power0.99, required n d. Now let's see what happens to these calculations if we assume the effect size is a little bit larger. Using d 0.6, re-do the calculations from part c. above: i. For power 0.55, required n ii. For power 0.8, required n ili. For power-0.99, required n e. Now suppose that the sample size available to you is fixed. In this case, the main interest is finding out what your power will be, given sample size Find the power of the test for each combination of effect size and total sample size below, and round your answers to three decimal places. Assuming effect size-0.5: i. When total n-6, power i. When total n-36, power ili. When total n-90, power For effect size 0.8: i, when total n = 6, power = ii. When total n 36, power ili. When total n 90, power f. Suppose it turns out that the true population effect size is much smaller than we thought, e.g. d 0.1. Also suppose we use a small sample size, e.g. n following should be true? Select all that apply. 20. which of the We will probably reject Ho We will probably fail to reject HO We will probably commit a Type II error In order to detect a "significant" difference, We will probably commit a Type I error In order to detect a "significant" difference, we will have to underestimate the true difference in mean litter sizes we wl have to overestimate the true difference in mean litter sizes

Answer 1

Cc) We knoa that 2. i) F 0Q67 분 О.SS p 0.13 Cu C)e 2 (o.5)1 x 576 (2.01) O 85 0x5 74x 4.368, O 8s So.3905 0、35

2 2 2 Co.s) x516 x (2.3) 0.25 2x5.16 x 7.8 0.85 90. 38 0.25 361 No n 2 d.5 2 x 5.76 x(4.21)

e 21202 0.8s Cr) 2 2 (o-6)

2x5-16 x 21) 3 0 36 50-320S 0.36 0.84 2 0.6 2 2 0.3168 0- 36 Q5 I e-n - Q,'3

5 2 2-33-f 2x5-76 x21)2 0.36 0.36 0.36 Cf) wẹ Gill f.noba bb fajt to eject NullhgR tesis (H.

Note:Due to insufficient  limit i couldn't solve e bit.If you want to answer for e bit please post it separately i will answer it.Thank you.