HW 8 Grading Key Due 11/29/05

8.1 Sampling designs.  See ../StatPrimer/independ-odd.htm
[3 pts]
8.2 Sampling designs, 2. 
(A) Needle stick injuries. Independent samples
(B) Public health laboratory. Single sample.
(C)  Public health laboratory, 2. Paired samples. 
8.5 Large t statistic. See ../StatPrimer/independ-odd.htm
8.7 Bone density in newborns. See ../StatPrimer/independ-odd.htm
8.8 Joggers. 
  • df1 = 25-1 = 24, df2  = 26-1 = 25, df = 24 + 25  = 49 
  • Use df = 50 as closest approximation and t50,.975 = 2.01
  • s2pooled = [(24)(4.82) + (25)(5.12)] / 49 = 24.60
  • sexbar1-xbar2 = [24.60 � (1/25 + 1/26)] = 1.389
  • 95% CI for �1 - �2 = (47.5 - 37.5) � (2.01)(1.389) = 10.0 � 2.8 = (7.2, 12.8)
  • Interpretation:  We are 95% confident that the population mean difference �1 - �2  lies between 7.2 and 12.8. (The source  http://www.thefitnessexpert.com/cardio_coach.html  suggests that this is quite a meaningful improvement in metabolism.)
8.9 Cytomegalovirus. See ../StatPrimer/independ-odd.htm
[7 pts]
8.10 Hemodialysis and anxiety. State-Trait Anxiety Inventory Score
  • H0: �1 = �2 vs. H1: �1 not = �2  
  • df1 = 38 - 1 = 37, df2  = 23 - 1 = 22, df = 37 + 22 = 59 
  • s2pooled = 96.277
  • sexbar1-xbar2 = 2.5922
  • tstat = (33.42 - 39.71) / 2.5922 = -2.43 with 59 df (use either df = 50 or df = 100 as closest approximation)
  • Two sided p value: .01 < p < .02
  • The difference is  significant. There does appear to be some benefit associated with the progressive relaxation tapes.
SS.1 Vegetarians and non-vegetarians. See ../StatPrimer/independ-odd.htm
SS.2. Allopathic and holistic medicine.Use n = 116 per group.