statistics

Ruth2014

Biostatistics_CaseStudy11.p.docx

Home >Science homework help >statistics

Biostatistics Case study:

of Association for Categorical Data

There are seven questions at the end of this article.

1. a. a and c

The type of research design described in the abstract is retrospective and observational design. This is because information about the individuals eligible for the study was collected on the basis of entry date into the TB registry and then the medical records were studied for 15 months after the confirmation of the PTB.

2. The proportion of patients who were successfully treated is 275/399 = 68.9%.

3. SE of p = sqrt of (pq)/n = sqrt (0.689 x 0.311)/399) = 0.02316956

95% CI for

Upper CI = 0.6892231+1.96*0.02316956 = 0.7346354 or 0.73

Lower CI = 0.6892231-1.96*0.02316956 = 0.6438108 or 0.64

4. d. a and c.

In other words, the 95% CI calculated above means that we are 95% certain or confident that the true population proportion of the patients who were successfully treated lies between 0.64 and 0.73

5. Observed Treatment Success

DOT	Yes	No	Total
Yes	186	64	250
No	86	60	146
Unknown	3	0	3
Total	275	124	399

DOT by Treatment Success

Count Expected	Yes	No	Total
Yes	172.306	77.6942	250
No	100.627	45.3734	146
Unknown	2.06767	0.93233	3
Total	275	124	399

6. Contingency Table

DOT By Treatment Success

Count

Total %

Col %

Row %

Yes

186

46.97

68.38

74.40

16.16

51.61

25.60

250

63.13

21.72

31.62

58.90

15.15

48.39

41.10

146

36.87

Total

272

68.69

124

31.31

396

Test	Chi-square	Prob>Chi-square
Pearson	10.290	0.0013
Fisher's Exact Test	Prob	Alternative Hypothesis
Left	0.9995	Prob(Treatment Success=no) is greater for DOT=yes than no
Right	0.0011	Prob(Treatment Success=no) is greater for DOT=no than yes
2-Tail	0.0016	Prob(Treatment Success=no) is different across DOT

Alpha = 0.05

df = __2___ (5 pts)

Critical value = __5.99___ (5 pts)

Chi-squared test statistic = __11.69587___ (5 pts)

a. Successful outcome is dependent on DOT status.

7. e. a, b, and c.