Basic Applied Statistics 200
Solutions to Midterm 1 at 10:00
-
- (iii) 5 number summary values (and boxplots) are higher for males
- (iii) Stdev is higher for males, and boxplot shows more spread
- (iii) there is some right skewness, and 3 high outliers
- (iii) for right skewness/high outliers, mean > median
- Five Number Summary is better in the case of skewness and/or outliers
-
- stems should be 0 (leaves 12223), 0 (leaves 556), 1 (leaves 00002),
1 (leaves 55), 2 (leaves 00)
- no: shape is neither symmetric nor bell-shaped
- 10 (the 9th value of 17)
- (ii) 6 [the numbers look to be about 6 away from their mean, typically]
-
- 1-.6700=.3300, or by symmetry look up z=-.44
- .0228-.0013=.0215
- -1.64 or -1.65
-
- 500 plus or minus 2(115): between 270 and 730
- x < 300 means z < -1.74; proportion is .0409
- the mean, 500
- cut-off for top 15% same as for bottom 85%: z= +1.04, so x=500+1.04(115)
=619.6
-
- heights
- (i) positive (iii) linear (v) strong
- positive square root of .623 (Rsquared) is .79
- .79 (r unaffected by change in units of measurement)
- -15.4 + 0.364(67) = 9
- 8 - 9 = -1
- (iii) [the output indicates both R and X for this observation]
- (ii) without that student, the points exhibit tighter clustering
- (i) a random residual plot indicates a straight line alone is a good
summary
-
- males: 13/20 = .65 vs 19/54 = .35
- 35/42 = .83
- 35/74 = .473
-
- (i) an observational study; imposing a treatment of meals would be
unrealistic
- (i)
- (iii)
-
- Pitt students constitute the population
- 3 times 2 = 6
-
- (ii) treatment is chewing, so control should be not chewing
- (i) randomness avoids bias
- (i) adequate replication is another principle of good design
[ Home
| Calendar
| Assignments
| Handouts
]
|