Basic Statistics
Chapter 2 Descriptive Statistics
2.1 Exercises
-
2.1.1 The data presented below refer to the yields of \(50\) plants of an \(F_2\) progeny of common bean in g/plant, evaluated at the Department of Biology of UFLA, in \(1997\).
2.81
3.19
3.49
3.76
6.02
8.23
2.23
3.01
4.43
13.94
3.10
1.52
3.38
2.85
4.64
7.33
6.78
13.12
13.84
9.40
6.20
2.39
9.19
7.07
9.20
13.46
3.90
8.99
7.97
5.15
12.95
25.52
6.61
16.56
9.60
6.71
6.73
3.86
3.50
4.80
8.40
13.86
6.53
18.44
22.14
9.15
8.75
10.86
14.20
10.09
-
a) Group the data into a frequency distribution table in several different ways, determining the number of classes by the criteria presented in Section 2.1.
-
b) Choose one of the criteria and draw the histogram and the frequency polygon on the same graph.
-
c) Construct the cumulative frequency distributions and draw the ogives on the same Cartesian plane.
-
d) What is the percentage of plants with yield greater than \(9\) g/plant? Use the ogives and algebraic interpolation from the frequency distribution. Compare and discuss the results with the exact sample proportion obtained from the elaborated data.
-
e) Discuss the nature of the distribution regarding symmetry based on the inspection of the constructed frequency polygon.
-
f) Obtain the stem-and-leaf representation, discuss and compare the results with the polygon and histograms obtained.
-
g) Above which yield are \(50\)% of the plants (\(25\) plants)? Use all available resources to obtain this value, compare and discuss the results obtained with each of them.
-
h) What is the percentage of plants with yield below 3.5 g?
-
i) Obtain the yield above which \(25\)% of the plants have higher yields and the yield below which \(25\)% of the plants have lower yields.
-
-
2.1.2 The data below refer to the number of nematode galls observed in \(85\) sampling units (plant roots) from a sample taken at UFLA, Lavras, MG.
Number of nematodes Frequencies - Plants 0
35
1
20
2
17
3
\(\,\,7\)
4
\(\,\,3\)
5
\(\,\,2\)
6
\(\,\,1\)
-
a) Draw a graph of the occurrence of nematode galls in the evaluated plants.
-
b) Obtain the cumulative distribution table and the corresponding relative frequencies.
-
c) Is it possible, in your opinion, to find a plant with more than \(6\) galls? Justify.
-
-
2.1.3 Let the following properties of summations hold. Consider \(a\), \(b\), and \(k\) as constants.
-
i. \(\displaystyle \sum _{j=1}^n aX_j = aX_1+aX_2+\cdots +aX_n=a\sum _{j=1}^n X_j\)
-
ii. \(\displaystyle \sum _{j=1}^n Y_jX_j = Y_1X_1+Y_2X_2+\cdots +Y_nX_n\ne \sum _{j=1}^n Y_j\sum _{j=1}^n X_j\)
-
iii. \(\displaystyle \sum _{j=1}^n (aX_j +bY_j)= a\sum _{j=1}^n X_j+b\sum _{j=1}^n Y_j\)
-
iv. \(\displaystyle \sum _{j=1}^n k = nk\)
-
v. \(\displaystyle \sum _{j=1}^n X_j^2 \ne \left (\sum _{j=1}^n X_j\right )^2\)
-
vi. \(\displaystyle \sum _{j=a}^b k =(b-a+1)k\), \(\qquad 1\le a\le b \le n\)
Consider also the example of variables arranged with double identification. For example, a researcher interested in verifying whether the mathematics grades of high school students differ among Brazilian states sampled schools and recorded the grade of \(200\) students in each state. To identify these data, the statistician could identify the states by one index and the student by another. Thus, we have the variable \(X\) with two indices (\(X_{ij}\)). The index \(i\) represents the Brazilian states and the index \(j\), the students of each of them. Another example refers to the average yield (tons) per hectare of a grass after the use of nitrogen and phosphate fertilizers. Three amounts of nitrogen N were applied and four doses of phosphorus P were also considered.
Yield in t/ha of a forage grass under the effect of \(3\) doses of \(N\) in combination with \(4\) doses of \(P\) observed in a zootechnical experiment.
Nitrogen content (\(j\)) Phosphorus content (\(i\))
1
2
3
1
4.6
5.0
5.5
2
5.0
5.5
6.1
3
5.2
5.8
6.4
4
6.0
6.2
6.8
In some statistical analyses it is often necessary to sum the rows and/or columns, as well as the entire table. The summation notation can be used for this purpose. Since two factors determine the yield, two indices are used to represent them, as commented above. Thus, two summation symbols can be used in some cases. Accordingly, the following summations are defined:
-
i. Sum all yields of the previous table.
\(\seteqnumber{0}{2.}{0}\)\begin{align*} \sum _{i=1}^4\sum _{j=1}^3 X_{ij} & = X_{11}+X_{12}+\cdots +X_{43}\\ &= 4.6+5.0+\cdots +6.8 = 68.1. \end{align*}
-
ii. Sum each of the rows
\(\seteqnumber{0}{2.}{0}\)\begin{align*} \sum _{j=1}^3 X_{ij} & = X_{i1}+X_{i2}+X_{i3}, \quad \forall i=1, 2, 3, 4. \end{align*}
Thus, for example, for the second phosphorus dose (\(i = 2\)), the total yield is
\(\seteqnumber{0}{2.}{0}\)\begin{align*} \sum _{j=1}^3 X_{2j} & = X_{21}+X_{22}+X_{23} =5.0+5.5+6.1 =16.6. \end{align*}
-
iii. Sum each of the columns
\(\seteqnumber{0}{2.}{0}\)\begin{align*} \sum _{i=1}^4 X_{ij} & = X_{1j}+X_{2j}+X_{3j}+X_{4j}, \quad \forall j=1, 2, 3. \end{align*}
Thus, for example, for the third nitrogen dose (\(j = 3\)), the total yield is
\(\seteqnumber{0}{2.}{0}\)\begin{align*} \sum _{i=1}^4 X_{i3} & = X_{13}+X_{23}+X_{33}+X_{43}=5.5+6.1+6.4+6.8=24.8. \end{align*}
With this information, solve the following exercises. Let the data sets be:
\(\qquad \qquad X=\{2, 4, 4, 3, 2\},\)
\(\qquad \qquad Y=\{1, 2, 3, 6, 7\}.\)
Obtain
i. \(\displaystyle \sum _{j=1}^4 X_j\) iv. \(\displaystyle \sum _{j=1}^5 Y_j\) ii. \(\displaystyle \sum _{j=1}^5 4X_j^2\) v. \(\displaystyle \sum _{j=1}^5 X_jY_j\) iii. \(\displaystyle \sum _{j=1}^5 (3X_j+2Y_j)\) vi. \(\displaystyle \sum _{j=1}^4 X_jY_j\) + \(\displaystyle \sum _{j=1}^5 Y_j^2\) Demonstrate numerically and algebraically that \(\displaystyle \sum _{j=1}^n(X_j-\bar {X})=0\). Use the data from set \(X\) to show it numerically.
Obtain, based on the grass yield table, the following sums
\(i. \displaystyle \sum _{i=1}^4\sum _{j=1}^3 X_{ij}^2\) \(\qquad \quad ii. \displaystyle \sum _{i=1}^4 X_{ij}^2 \quad j=1,2,3\) \(\qquad \quad iii. \displaystyle \sum _{j=1}^3 X_{ij}^2 \quad i=1,2,3,4\)
-
-
2.1.4 A survey of the daily milk yield of \(30\) rural producers served by the “Panela Cheia” program was carried out in the western region of Paraná, in the municipality of Marechal Cândido Rondon, in \(1992\) (Roesler 1997). The results of the daily yield of the \(30\) producers are presented below.
8.13 8.23 8.60 8.80 8.97 9.05 9.12 9.30 9.35 9.78 9.80 9.86 9.90 9.95 10.00 10.11 10.13 10.15 10.16 10.23 10.31 10.33 10.40 10.46 10.50 11.14 11.29 11.46 12.05 12.14 -
a) Obtain the estimate of the arithmetic mean and the midrange (\(\bar {X}_{0.50}\)).
-
b) Estimate the trimmed mean, removing the two largest and two smallest values from the sample. In this case, do you think the use of the trimmed mean is justified?
-
c) Estimate the median. Do you think it would be possible to estimate the mode in these data as they are?
-
d) Using the \(5\) methods presented (equations (2.2.8) to (2.2.12)), estimate the \(5\%\) and \(95\%\) percentiles. Did the methods show many differences among themselves?
-
e) If the yield of each producer is multiplied by \(R\$\,\,0.27\) to obtain the average income per producer/animal/day, what will be the new value for the sample?
-
f) Estimate the harmonic mean and the geometric mean and compare the values among themselves and also with the arithmetic mean.
-
-
2.1.5 Make the frequency distribution of the data from Exercise 2.1.4 and calculate:
-
a) Arithmetic mean, median, and mode.
-
b) The geometric mean, the harmonic mean, and the quadratic mean.
-
c) Compare these values with those obtained in the previous exercise and discuss the reasons for the differences.
-
d) Draw the histogram and the frequency polygon.
-
e) Based on these graphs, determine the nature of the distribution regarding symmetry. Based on this answer, indicate which measure of location is most appropriate to represent the sample data. Justify.
-
f) If you were asked by the city mayor to estimate the total daily milk yield of the city, how would you do it? Additional information - number of milk producers in the city: \(7.309\); Average number of cows per producer (sample mean): 11.80 cows/producer; Average number of lactating cows: 8.075.
-
-
2.1.6 Use the data from Exercise 2.1.2 and estimate the mean, mode, and median.
-
2.1.7 A survey of zootechnical indices was carried out in the western region of Paraná, in the municipality of Marechal Cândido Rondon, in \(1992\), sampling \(20\) rural producers served by the “Panela Cheia” program (Roesler 1997). The results of the calving intervals (in months) of the \(20\) producers are presented below.
11.80 11.90 12.00 12.30 12.80 12.99 13.10 13.50 13.80 14.10 14.55 14.65 14.70 15.00 15.10 15.20 15.50 15.80 15.90 15.96 -
a) Estimate the total range, variance, standard deviation, and mean deviation.
-
b) Obtain the coefficient of variation.
-
c) Obtain the standard error of the mean and describe how you differentiate what the standard deviation measures from what the standard error of the mean measures.
-
d) Obtain the coefficient of precision of the mean for these data and for the milk yield data (Exercise 2.1.4) and verify in which case the mean was estimated with more precision.
-
e) In each previous case, comment on the meaning of the estimate obtained and on how they should be applied.
-
f) If each data point is divided by \(12\) to obtain the calving interval in years, what are the new values of the range, variance, standard deviation, coefficient of variation, mean deviation, and standard error of the mean?
-
g) Verify whether Chebyshev’s theorem was satisfied in this example, given \(P(\bar {X}-kS<X_i<\bar {X}+kS)\ge 1-\dfrac {1}{k^2}\), for \(k=1\), \(2\), and \(5\).
-
-
2.1.8 Make the frequency distribution of the data from Exercise 2.1.7 and calculate:
-
a) Range, variance, standard deviation, coefficient of variation, and standard error of the mean?
-
b) Compare these values with those obtained in the previous exercise and discuss the reasons for the possible differences observed.
-
c) If you were asked to represent the data by two measures, which measures would you use and why?
-
d) After the Panela Cheia program, the calving interval showed a mean of 13.85 and a standard deviation of 2.00 months. Which situation showed greater variability, before or after the program? Justify your answer.
-
-
2.1.9 Demonstrate the following properties of the variance:
-
i. Let \(Y_i = X_i + k\), where k is a constant. Let \(\bar {X}\) be the mean and \(S^2_X\) the sample variance of the \(n\) values of \(X\). Show that \(S^2_Y = S^2_X\).
-
iii. Demonstrate what the estimators of the range, mean deviation, standard deviation, standard error of the mean, coefficient of precision, and coefficient of variation of the variables \(Y\) and \(Z\) from items 2.1.9i and 2.1.9ii become.
-
-
2.1.10 Obtain the estimates of the total range, variance, standard deviation, coefficient of variation, and standard error of the mean for the nematode gall data (Exercise 2.1.2).
-
2.1.11 Estimate the coefficient of skewness and kurtosis of the milk yield data from the sample of \(30\) producers of Marechal Cândido Rondon and discuss the results found.
-
2.1.12 Below are estimates of the coefficient of skewness and kurtosis in several samples. Classify them regarding symmetry and degree of flatness of the frequency distribution.
Coefficient of Coefficient of Classification of Classification of the degree skewness kurtosis symmetry of flatness +0.5 3.0 …
-2.0 1.0 …
+2.0 2.0 …
+3.0 3.0 …
+0.0 3.0 …
+0.0 3.5 …
-3.0 4.5