You are on page 1of 15
This section covers the computation of the midhinge, interquartile range and the quartile deviation which are extension of the measures of dispersion and location. A. Midhinge The midhinge is the mean of the first and third quartiles in the data set. It is used to overcome potential problems introduced by extreme values (or outliers) in the data set. It 1s computed using Formula 4-18. Midhinge= 222 (Formula 4-18) ge of the ages of 9 middlemanagement employees of a certain Example J: Find the midhin, 59, 48, 54, 46, 51, 58, and 55. company. The ages are 53, 45, Solution: ~ . Take note that we already computed the values of Q) ~47 and Qs = 56.5 of the data in Section 4.5, Example 1. We can now apply Formula 4-18 to determine the value of the midhinge. 474965 _ 1035 , Midhinges S22 _ M1888 ES oor Thas, the midhinge is 51.75- B. Interquartile Range the interquartile range UQR), also called the mnidspread or ical dispersion, being equal to the difference between, the In descriptive statistics, middle fifty. is 3 measure of statist third and first quartiles. i statistic, having a breakdow, Unlike the range, the inert ange et isatso used to bulld box gj” the total 1 . 25%, and is thus often preferzed to CFoemals 419 Interquartile Range (IQR) = Qs- i it Example 2: Find the inerquartile range of the ages of 9 mile menagemen employees 4 certain company. The ages are $3, 45, $9, 48, 54, 46, 51, 58, ar Solution: Remember that Qi = 47 and Qs = 56.5, We can now make use of Formula 4-19 to determine the interquartile range. Interquartile Range (IQR)= Q)=.0) = 565-47 = 95 Hence, the interquartile range is 95. ©. Quartile Deviation The quartile deviation is a slightly better measure of absolute dispersion than the range, By it ignores the observation on the tails. If we take different samples from a population and calculate their quartile deviations, their vaiues are quite likely to be sufficiently different. This's called saimpling fluctuation, It ig not a populor measure of dispersion, The quartile deviate cakulated from the sample dala does not help us to draw any conclusion (inference) abaut te quartile deviation in the population, ‘ Quartile Deviation (QD} = 58 , _ Formula 4-20) . : Example 3: Find the quartile deviation of the ages of 9 middiemanagement employees af 3 , certain company, The ages are 53, 45, $9, 48, 54, 46, 51, 58, and 55. . Solution: Remember that Ch = 47 and Qsw 56.5. - , We can now make use of Formula 4-20 to determine the quartite deviation, Quartile Deviation (en) S52 5-47 Hence, the quartile deviation Is 4.73, 25.475 2. “t Example 4: Using the example provided in Chspter 2 on SIS Travel Agency, Dg ine the micdhinge, interquartile range, and quartile deviation of the frequency distribution on th 50 people taking travel tours. ages of Solotion: Since we already obtained the values for Qi = 40 and Qs = 58.62, we can simply substitute itto our formulas to determine the value of mi ge. interquartile range, and quartil deviation, We can now make use of Formula 4-18, Formula 4-19, and Formala 4-20 to delermine the Values of midhinge, interquartile range, and quartile deviation, respectively. +0, 4045882 9882” Midhinge= = : $8? Laoat Interquartile Range QR) =Qr—Q) =58.82- 40 = 18.82 Quartile Deviation (py = S35. SBAR—A 1482 =941 Hence, the midhinge is 49.41, the IQR is 18,82, and the QD is 9.41, in any given two samples with the same units of measures, the variance and standard deviation for each can be compared. In cases when one is interested to compare standard deviations of two different units, coefficient of variations can be applied. The coefficient of variation, denoted by CV, is the standard deviation divided by the mean. The sesult is expressed aS a percentage. . For Sample: Cv = S(100%) 7 {Formula 4-21) ForPopulation: = cy == 100%) ‘ (Formula 4-22) 2 : where: CV = coefficient of variation. . X = sample mean. 3™ sample standard deviation. = population mean. = population standard deviation. Example 1: The average age of the engineers at VSAS Pipeline Corporation is 20 years, with a standard deviation of 3; the average monthly salary of the ‘engineers is P45,000, with standard deviation of P3150. Determine the coefficient of variations of age and salary. Solution: : Collect the information needed to compute for the values of coefficient of variations 3-3, X,= 30, s1= 13,150, R= Pas 000 Compute for the eaefficient of variations of age and salary. - 5, 3 CV, = 21 (109%) = — (100%) = 10% age xX 00 3 = £1. 00%) = 229_o0%) = 7%" salary CV, = 100%) = Fe gy 100) = 7% Y Since the coefficient of varii is larger for age, the ages are more variable than the salary, ver L-year period is P12,500, and Example 2;"The mean of commissions of Educational Insurance 0 the variance is : the standard deviation is P1,350, The mean of the number of sales is 69, and Compare the variations of the twa. . Solution: . Collect the information needed lo compute for the values of coefficient of variations ° zo a= P1350, 712500, gz = 56.25, x ‘Compute for the coefficient of variations on commissions and sales. The coeificients of variations aré * . “oma * cy, = £10004) = 22% (0m) = 108% commission 1 2,500 73 33, 212.5% sales oy, 75 100%) 12.5% = 22 (100%)= REE Xy a Take note that we have to extract the square root 56.25 to obtain the standard deviation onsales variance. : Since the coefficient of variation is larger for sales, the sales are mort variable than the commissions. Ao already mentioned, the variance and standard deviation of a variable can be used to determine the dispersion, or spread, of a variable. Specifically, the larger the variance and standard devietion, the more the data values are spread or dispersed. The fusiin mathematician P. L. Chebychev (1891-1694) developed a theorem that specifies the proportions of the spread in terms of the standard deviation. . A. Chebysheo's Thearems, For any set of observations, the proportion of the values that ie wit kstandard deviations of the mean is at least ie e where k is anly constant greater than! Example 1: The anean price of tptop computer is P25,500 and the standard deviation is P257- Find the price range for which at least 88.89% of the laptop will sell, leviation 3 " Solution: Chebychev’s theorem states that 88.89%, of the data values mean. Hence, . P25,500 + 3(P2,500) ~ F25,500 + P7500 = F33,000 : p2s.s00 -3(P2.500) = P25.500-P7,500= 18,000 . Therefore, at least 88.89% of ali laplop sold will have a price range from P18,000 and P33,000. fall within 3 standards of Example 2; A Survey conducted by Commission on Higher Education (CHED) found that the mean amount of taining allowance for department heads of colleges and universities was 25,000. The standard ‘deviation was P1500. Using Chebychev’s theorem, find the minimum Percentage of the data values that will falf between 122,000 and P28,000, Solution: Oép 7: Subtract the mean from the larger value. 28,000 - P25,000 = P3,000 Sty & Divide te difference by the standard deviation to obtain k. Therefore, at least 75% of the data value will fall between P22,000 and P28,000. 5. Empirical Rule. For a symmetrical, bell-shaped frequency distribution, 1, About 68% of the area under the normal distribution curve is within one standard deviation of the mean, This can be written as jst 10. 2 About 95% of the area under the normal distribution curve is within two standard deviation of the mean. This can be written as p20. 3. About 99.7% of the area under the normal distribution curve is within three standard + deviation of the mean. This can be written as ys + Jo. p-3e H- a Bb ptla p+20 pro a) \ oa = J 7% ing. In stal lo wtos or kurtos, meaning, om the Greek word RYT One the distribution of observed dala 20rd th 4] measure used to ribution (as compared to y, or flatness of a dist ° u ene ‘A kurtosis of a data set is computed using Kartosis it excess) is a stati mean, It measures the relative normal distribution, which shows & kurlosis of zero). the forrula: nin+ 1) x,-*)]|__ ae (Formula 4-23) bute late 2) ( $ ] (n-2}n-3) where: kurt © kurtosis. n= sample population, X= the value of any particular observations: X =sample mean. s= sample standard deviation. or measurement, Three Types of Kurtosis. 1. Leptokcurtic are distributions where values chistered heavily or pile up in the center, There are tall distribution with narrow humps and long and high fails. its kurtosis Is positive (kurtosis > 0) and it denotes a high degree of peakedness. 2. Mesokurtic are intermediate distribution which are neither too. peaked por too fiat, The values are immediately distributed about the center. Its kurtosis is zero (kurtosis= 0). . 3. Platykurtie are flat distributions with values more evenly distributed about the center with broad humps and shot tails. Its kurtosis is negative (kurtosis

You might also like