Question

1. In 2004, Murthy, Krumholz, & Gross investigated disparities in access to clinical trials in an article entitled “Participation in cancer clinical trials: race-, sex-, and age-based disparities”.

Race/Ethnicity	White, non-Hispanic	Hispanic	Black	Asian/Pacific Islander	American Indian/Alaskan Native
Population Distribution	75.6%	9.1%	10.8%	3.8%	0.7%
Participants in Clinical Trials	3,855	60	316	54	12

Does the distribution of participants in clinical trials correspond to the population distribution, given that…

a.

b.

3. What would you conclude based on these findings?

Answer 1

null hypothesis:Ho:distribution Distribution in trails is same as distribution in population.

Alternate hypothesis:Ha:distribution Distribution in trails is not same as distribution in population.

degree of freedom =categories-1=

4

for 0.05 level and 4 df :crtiical value X2 =

9.4877

for 0.01 level and 4 df :crtiical value X2 =

13.2767

applying chi square goodness of fit test:

	relative	observed	Expected	residual	Chi square
category	frequency(p)	Oi	Ei=total*p	R2i=(Oi-Ei)/√Ei	R2i=(Oi-Ei)2/Ei
1	0.756	3855.000	3248.53	10.64	113.221
2	0.091	60.000	391.03	-16.74	280.234
3	0.108	316.000	464.08	-6.87	47.248
4	0.038	54.000	163.29	-8.55	73.144
5	0.007	12.000	30.08	-3.30	10.866
total	1.000	4297	4297		524.7132
test statistic X2 =		524.713

a)

since test statistic is higher than critical value at 0.05 level we reject null and conclude that  distribution of participants in clinical trials differs from to the population distribution,

b)

since test statistic is higher than critical value at 0.01 level we reject null and conclude that  distribution of participants in clinical trials differs from to the population distribution,

c)

we have sufficient evidence to conclude that distribution is different and specially Hispanic race has highest difference,