MATH 1040 Statistics for Laboratory Sciences
Review Questions Answers
Question 1
Self-selected sample; matched pairsH0: md = 0; H1: md < 0 xss=removed>P-value = 013 < 0>H0; the oat bran diet is effective in reducing cholesterol
Question 2
Control sample cant be assumed to have come from Normal distribution; use Kruskal-Wallis
H0: All medians are equal; H1: Not all medians are equal; a= 0.05
P-value = 0.116 > 0.05 so cant reject H0
Question 3
H0: Nutritional status and illness are independent; H1: Nutritional status and illness are dependent; a =05.
All expected counts are ?3;5; P-value = 000 < 0>H0.
g. no illness and normal diet status shows higher than expected count
Question 4
H0: p1 p2 = 0; H1: p1 p2 ?1;0; a= 0.05
P-value = 0.002 < 0>H0. Difference in proportions is statistically significant at 5% level. Surgery appears to be better.
Question 5
Check Normality and standard deviations
H0: All means are equal; H1: Not all means are equal; a= 05
P-value = 229 > 0.05 so cant reject H0
Specific age range, specific location
Question 6
Two independent samples, both come from Normal populations so use two-sample t-test
H0: m1 m2 = 0; H1: m1 m2 < 0 a=05>P-value = 019 < 0>H0; men score lower than women on average.
Question 7
Odd ratio = 11.85
Question 8
P-value = 000 < 0>Average rate of increase in systolic blood pressure with age46 0.2314P-value = 000 < 0>Excellent fit; 8% of variability in systolic explained by age95% PI is (151.119, 165.291). Prediction accurate since (i) R2 indicates excellent fit; (ii) Margin of error is small relative to fitted value; (iii) Prediction is by interpolation
Question 9
H0: Median = 0; H1: Median ?1;0; a= 0.10
P-value = 0.1797 > 0.10 so cant reject H0; not enough evidence to conclude there is a difference in hunger rating.
Question 10
Check Normal probability plots and standard deviations
H0: All means are equal; H1: Not all means are equal; a= 0.05; P-value = 0.02 < 0>H0.
None (diet with no added fat) is different from the others; no significant difference between the two diets with added fat.
All aged 12, from the same
Question 11
0.3783, so low study power.
Question 12
YesOn average, metabolic rate increases by 25.19 calories for every extra kg in lean body mass. On average, the metabolic rate is expected to be 187 calories with 0 kg in lean body mass (not possible). Lean body mass explains 75.2% of variability in metabolic rate.Some evidence of unequalSlope estimate is significant, P-value < 05>Intercept estimate is not significant, P-value > 0.05. It would make sense to fit a no- intercept model.Use confidence Prediction should be reasonably accurate