INDE 315 Homework #4 Name (Last, First)_________________________________
Fall 2012 Student ID#_________________________________
Using the data on middle finger mass (titleled MiddleFingerMass (gr)) from the 132 participants from the INDE 315 data set answer the following questions.
1. Can data collected on middle finger mass from the 132 participants be considered a random sample representative of the University of Washington student population? Answer YES or NO and justify your answer. (4 points). 2 points for NO, 2 points for correct explanation
NO. The population of the University of Washington has staff and graduate students not represented in this sample. In addition there is likely a greater proportion of males than in the university populations
2. Can data collected on middle finger mass from the 132 participants be considered a random sample representative of the University of Washington undergraduate engineering student population? Answer YES or NO and justify your answer. (4 points) 2 points for NO, 2 points for correct explanation
NO. The population of the University of Washington engineering undergraduates has other races besides Asian and Caucasian students. In addition, freshman and sophomore students may be underrepresented in the sample
YES. The does represent the University of Washington engineering undergraduate student population since all engineering majors take this class and the proportion of males and females is likely similar to the overall engineering population.
3. Calculate the following for the middle finger mass data (hint: using Excel will be a big help here). You can paste a screen snapshot of your excel output if you like. (24 points) (2 points for each correct answer, they dont need to label all values as grams to get credit a) Sample mean, 20.81 gr b) Sample variance, 40.09 gr 2
c) Sample standard deviation, 6.33 gr d) Sample standard error, 0.55 gr e) Median, 20.4 gr f) 1 st Quartile Value, ~16.4 gr (strange my JMP gives slightly different values, see below) g) 3rd Quartile Value, ~ 25.0 gr (strange my JMP gives slightly different values, , see below) h) Beginning and ending values of the Interquartile Range (IQR) and the size of the IQR Beginning 16.4, Ending ~ 25.0 IQR i) The value 1.5 IQR below the 1 st quartile, 12.9 gr or 3.5 gr to 16.4 gr j) The value 1.5 IQR above the 3 rd quartile 12.9 gr or 25.0 to 37.9 gr k) Smallest data point which makes the lower end of the box and whisker plot, 8.8 gr l) Largest data point which makes the upper end of the box and whisker plot, 37.1 gr
4. Draw the box and whisker plot for the middle finger mass data and label and provide the associated middle finger mass values for all the important points on the plot. (note: you can draw the box and whisker plot by hand if you prefer). (12 points)
(2 points for box plot, 2 points for each correctly labeled item (lower whisker, upper whisker, 1 st Q, median and 3 rd Q))
5. Construct a normal probability plot of the middle finger mass data. (hint: using Excel will be a big help here). Does it seem reasonable to infer that the data is normally distributed? Answer YES or NO and justify your answer. (9 points) 6 points for plot, 3 points for correct answer on normality, the percentiles can be on either the x or y axis No the subjects data does not appear to be normally since all the points dont fall on the line
J MP Says the data is not normally distributed, but just barely non-normal
6. Calculate the following for the middle finger mass data for the 49 Asian subjects and then repeat for the 83 Caucasian subjects (hint: using Excel will be a big help here). You can paste a screen snapshot of your excel output if you like. For the 49 Asian subjects (6 points) (2 points each answer, they do not need to have the label gr) a) Sample mean, 18.48 gr b) Sample standard deviation, 5.91 gr c) Median, 17.7 gr
For the 83 Caucasian subjects (6 points) (2 points each answer, they do not need to have the label gr) d) Sample mean, 22.19 gr e) Sample standard deviation, , 6.20 gr f) Median, 21.6 gr
7. Construct a normal probability plot of the middle finger mass data for the 49 Asian subjects and then construct a normal probability plot of the middle finger mass data for the 83 Caucasion subjects. (hint: using Excel will be a big help here). You can paste a screen snapshot of your excel chart if you like. Does it seem reasonable to infer that the Asian subject data is normally distributed. Answer YES or NO and justify your answer. Does it seem reasonable to infer that the Caucasian subject data is normally distributed. Answer YES or NO and justify your answer. (18 points)
6 points for plot, 3 points for correct answer on normality, the percentiles can be on either the x or y axis
YES the Asian subjects data appears to be normally as all except one point appears to be an outlier
8. The middle finger mass of the 132 subjects in the INDE 315 class is known to be normally distributed. You can calculate the sample variance from the data and assume the variance is known.
a) Construct a 95% two-sided confidence interval on the mean finger mass for the 83 Caucasian subjects (12 points)
x = 22.19, Standard Dev = 0.68, Z /2 = Z 0.025 = 1.96 Lower limit = x - Z /2 (/n) = 22.19 (1.96 6.2/83) = 20.84 Upper limit = x + Z /2 (/n) = 22.19 + (1.96 6.2/83) = 23.55
9. The middle finger mass of the 65 Caucasian males in the INDE 315 class is known to be normally distributed. You can calculate the sample variance from the data and assume the variance is known.
a) Construct a 95% one-sided, upper-confidence bound on the mean finger mass for the 65 Caucasian males. (6 points)
x = 23.79, Standard Dev of Mean = 0.71, Z
= Z 0.05 = 1.64 Upper Confidence Bound = x + Z
(/n) = 23.79 + (1.64 5.73/5) = 24.95 gr
b) Construct a 99% one-sided, upper-confidence bound on the mean finger mass for the 65 Caucasian males. (6 points)
x = 23.79, Standard Dev of Mean = 0.71, Z
= Z 0.01 = 2.33 Upper Confidence Bound = x + Z
(/n) = 23.79 + (2.33 5.73/5) = 25.44 gr
10. The middle finger mass of the 14 Asian females in the INDE 315 class is thought to be normally distributed. You can calculate the sample variance from the data but due to the small sample size it is unknown if this variance is representative (in other terms variance is unknown).
a) Construct a 95% one-sided, lower-confidence bound on the mean finger mass for the 18 Caucasian females. (6 points)
x = 14.47, Standard Dev of Mean = 1.07, t , n-1 = t 0.05, 13 = 1.77 Lower Confidence Bound = x + t , n-1 (/n) = 14.47- (1.77 4.02/14) = 12.57 gr
b) Construct a 90% one-sided, lower-confidence bound on the mean finger mass for the 18 Caucasian females. (6 points)
x = 14.47, Standard Dev of Mean = 1.07, t , n-1 = t 0.10, 13 = 1.35 Lower Confidence Bound = x + t , n-1 (/n) = 14.47 -(1.35 4.02/14) = 13.02 gr
11. How does the 95% lower-confidence bound for the 14 Asian females and the 95% upper-confidence bound for the 56 Caucasian males compare to the 95% two-sided confidence interval on the mean finger mass for the 132 subjects in the INDE 315 class? SIMILAR or DIFFERENT. (6 points). 14 Asian female 95 65 Caucasian male Lower Confidence Bound Upper Confidence Bound 16.44 gr 24.95 gr
132 Subjects Lower Limit Upper Limit 19.72 gr 21.9 gr (3 points)
Different, the CI on the 132 participants is smaller than the lower bound of the Asian females and the upper bound of the Caucasian males ( 3 points)