Skip to the page
Home Campus
Start free

Standard Deviation of Grouped Data

The standard deviation from a frequency table and from grouped data with mid-values, by both forms of the formula and by the calculator's statistics mode with a frequency column; why the answer for grouped data is an estimate.

What You Should Already Know

  • From Secondary 2: the mean of a frequency table is xˉ=ΣfxΣf\bar{x} = \dfrac{\Sigma fx}{\Sigma f}. For grouped data, xx is the mid-value of each class, and the answer is an estimate, because the values in a class are not known. For example, in the class 10≤t<2010 \le t < 20 the mid-value is 10+202=15\dfrac{10 + 20}{2} = 15.
  • From lesson 3.5: the standard deviation of a list of nn values is σ=Σ(x−xˉ)2n=Σx2n−xˉ2,\sigma = \sqrt{\dfrac{\Sigma(x - \bar{x})^2}{n}} = \sqrt{\dfrac{\Sigma x^2}{n} - \bar{x}^2}\text{,} and on a calculator it is σx, not sx.

Standard Deviation from a Frequency Table

In a frequency table, each value xx occurs ff times. So in the formulas of lesson 3.5 each squared deviation (x−xˉ)2(x - \bar{x})^2 is counted ff times, and so is each x2x^2; and the number of values nn is Σf\Sigma f.

The standard deviation of a frequency table

xˉ=ΣfxΣf\bar{x} = \dfrac{\Sigma fx}{\Sigma f}

σ=Σf(x−xˉ)2Σf\sigma = \sqrt{\dfrac{\Sigma f(x - \bar{x})^2}{\Sigma f}} or σ=Σfx2Σf−xˉ2\sigma = \sqrt{\dfrac{\Sigma fx^2}{\Sigma f} - \bar{x}^2}

fx2fx^2 means f×x2f \times x^2: square the value, then multiply by its frequency. It is not (fx)2(fx)^2.

Divide by Σf\Sigma f, the number of values, not by the number of rows of the table. The two forms give the same answer.

Example 1: The first form

The table shows the number of children, xx, in each of 40 families in a block of flats. Find the mean and the standard deviation of the number of children.

Number of children, xxNumber of families
0055
1188
221313
331010
4444
Total4040

Solution

Σfx=0+8+26+30+16=80,\Sigma fx = 0 + 8 + 26 + 30 + 16 = 80\text{,} so xˉ=ΣfxΣf=8040=2.\bar{x} = \dfrac{\Sigma fx}{\Sigma f} = \dfrac{80}{40} = 2\text{.} Each deviation is then a whole number.

Add a column for the deviation x−xˉx - \bar{x} and one for f(x−xˉ)2f(x - \bar{x})^2, the frequency times the square of the deviation: for x=0x = 0, 5×(−2)2=205 \times (-2)^2 = 20.

xxffx−xˉx - \bar{x}f(x−xˉ)2f(x - \bar{x})^2
0055−2-25×4=205 \times 4 = 20
1188−1-18×1=88 \times 1 = 8
2213130013×0=013 \times 0 = 0
3310101110×1=1010 \times 1 = 10
4444224×4=164 \times 4 = 16
Total40405454

Σf(x−xˉ)2=20+8+0+10+16=54.\Sigma f(x - \bar{x})^2 = 20 + 8 + 0 + 10 + 16 = 54\text{.}

σ=5440=1.35=1.161…\sigma = \sqrt{\dfrac{54}{40}} = \sqrt{1.35} = 1.161\ldots

The mean is 2 children and the standard deviation about 1.16 children.

Example 2: The second form

Find the standard deviation of the 40 families of Example 1 by the second form of the formula.

Solution

Add columns for fxfx and fx2fx^2. Each fx2fx^2 is f×x2f \times x^2: for x=3x = 3, 10×9=9010 \times 9 = 90.

xxfffxfxfx2fx^2
00550×5=00 \times 5 = 05×02=05 \times 0^2 = 0
11881×8=81 \times 8 = 88×12=88 \times 1^2 = 8
2213132×13=262 \times 13 = 2613×22=5213 \times 2^2 = 52
3310103×10=303 \times 10 = 3010×32=9010 \times 3^2 = 90
44444×4=164 \times 4 = 164×42=644 \times 4^2 = 64
Total40408080214214

σ=21440−22=5.35−4=1.35=1.161…\sigma = \sqrt{\dfrac{214}{40} - 2^2} = \sqrt{5.35 - 4} = \sqrt{1.35} = 1.161\ldots

σ≈1.16\sigma \approx 1.16, the same answer as Example 1.

Example 3: A mean that is not a whole number

The table shows the number of goals a school team scored in each of its 20 matches. Find the mean and the standard deviation of the number of goals.

Goals, xxMatches
0033
1166
2266
3344
4411
Total2020

Solution

The second form avoids decimal deviations.

xxfffxfxfx2fx^2
00330×3=00 \times 3 = 03×02=03 \times 0^2 = 0
11661×6=61 \times 6 = 66×12=66 \times 1^2 = 6
22662×6=122 \times 6 = 126×22=246 \times 2^2 = 24
33443×4=123 \times 4 = 124×32=364 \times 3^2 = 36
44114×1=44 \times 1 = 41×42=161 \times 4^2 = 16
Total202034348282

xˉ=3420=1.7\bar{x} = \dfrac{34}{20} = 1.7

σ=8220−1.72=4.1−2.89=1.21=1.1\sigma = \sqrt{\dfrac{82}{20} - 1.7^2} = \sqrt{4.1 - 2.89} = \sqrt{1.21} = 1.1

Mean 1.7 goals; standard deviation 1.1 goals.

Grouped Data

A grouped frequency table gives how many values lie in each class, but not the values themselves. As for the mean in Secondary 2, each value is taken to be at the mid-value of its class. Then the table is treated as a frequency table of the mid-values, and the same formulas give an estimate of the standard deviation.

The standard deviation of grouped data

Take xx as the mid-value of each class and ff as its frequency. Then

xˉ≈ΣfxΣf\bar{x} \approx \dfrac{\Sigma fx}{\Sigma f} and σ≈Σfx2Σf−xˉ2\sigma \approx \sqrt{\dfrac{\Sigma fx^2}{\Sigma f} - \bar{x}^2} (or the first form).

Both answers are estimates: the values in a class are spread out across it, not all at its middle, so the true mean and standard deviation may differ a little. Say so in the answer.

Example 4: Waiting times

The table shows the time, tt minutes, that 20 patients waited at a polyclinic one morning. Estimate the mean and the standard deviation of the waiting times.

Time, tt minutesPatients
0≤t<100 \le t < 1022
10≤t<2010 \le t < 2055
20≤t<3020 \le t < 3088
30≤t<4030 \le t < 4044
40≤t<5040 \le t < 5011
Total2020

Solution

The mid-values are 0+102=5\dfrac{0 + 10}{2} = 5, then 15, 25, 35 and 45.

Time, tt minffxx
0≤t<100 \le t < 102255
10≤t<2010 \le t < 20551515
20≤t<3020 \le t < 30882525
30≤t<4030 \le t < 40443535
40≤t<5040 \le t < 50114545
Total2020

Now work with the mid-values as a frequency table:

xxfffxfxfx2fx^2
55225×2=105 \times 2 = 102×52=502 \times 5^2 = 50
15155515×5=7515 \times 5 = 755×152=11255 \times 15^2 = 1125
25258825×8=20025 \times 8 = 2008×252=50008 \times 25^2 = 5000
35354435×4=14035 \times 4 = 1404×352=49004 \times 35^2 = 4900
45451145×1=4545 \times 1 = 451×452=20251 \times 45^2 = 2025
Total20204704701310013100

xˉ≈47020=23.5\bar{x} \approx \dfrac{470}{20} = 23.5 minutes.

σ≈13 10020−23.52=655−552.25=102.75=10.13…\sigma \approx \sqrt{\dfrac{13\,100}{20} - 23.5^2} = \sqrt{655 - 552.25} = \sqrt{102.75} = 10.13\ldots minutes.

Check by the first form: the deviations from 23.5 are −18.5-18.5, −8.5-8.5, 1.5, 11.5 and 21.5.

xxfff(x−xˉ)2f(x - \bar{x})^2
55222×342.25=684.52 \times 342.25 = 684.5
1515555×72.25=361.255 \times 72.25 = 361.25
2525888×2.25=188 \times 2.25 = 18
3535444×132.25=5294 \times 132.25 = 529
4545111×462.25=462.251 \times 462.25 = 462.25
Total202020552055

σ≈205520=102.75,\sigma \approx \sqrt{\dfrac{2055}{20}} = \sqrt{102.75}\text{,} the same.

The mean waiting time is about 23.5 minutes and the standard deviation about 10.1 minutes (estimates).

Example 5: The estimate and the true value

The times, in seconds, that 20 pupils took to run 200 m (lesson 12.10 of Secondary 2) are 37, 29, 42, 33, 35, 47, 31, 38, 40, 34, 28, 36, 45, 33, 41, 37, 32, 43, 35, 39. (a) Find the standard deviation of the 20 times. (b) Group the times in the classes 25≤t<3025 \le t < 30, 30≤t<3530 \le t < 35, …, 45≤t<5045 \le t < 50 and estimate the standard deviation from the table. (c) Compare the two answers.

Solution

(a) By calculator, keying in the 20 times: n=20n = 20, xˉ=36.75\bar{x} = 36.75, σx =5.048…= 5.048\ldots

(b) Count the times in each class, and use the mid-values:

Time, tt sffxx
25≤t<3025 \le t < 302227.527.5
30≤t<3530 \le t < 355532.532.5
35≤t<4035 \le t < 407737.537.5
40≤t<4540 \le t < 454442.542.5
45≤t<5045 \le t < 502247.547.5
Total2020

Key the five mid-values into the calculator with the frequencies 2, 5, 7, 4, 2: n=20n = 20, Σx=745\Sigma x = 745 and Σx2=28 375\Sigma x^2 = 28\,375.

xˉ≈74520=37.25\bar{x} \approx \dfrac{745}{20} = 37.25 and σ≈28 37520−37.252=1418.75−1387.5625=31.1875=5.584…\sigma \approx \sqrt{\dfrac{28\,375}{20} - 37.25^2} = \sqrt{1418.75 - 1387.5625} = \sqrt{31.1875} = 5.584\ldots

(c) The estimate, 5.58 s, is close to the true value, 5.05 s, but not equal to it. The table treats the 7 times from 35 to 39 s as seven times of 37.5 s, and so on: the single values are lost.

(a) σ≈5.05\sigma \approx 5.05 s (b) σ≈5.58\sigma \approx 5.58 s (an estimate) (c) The estimate is close to the true value but not equal, because each time was replaced by its class's mid-value.

Using a Calculator with a Frequency Column

For a table with many rows, use the calculator's statistics mode with its frequency column switched on. Each value (or mid-value) is then keyed in once, with its frequency beside it.

A frequency table on a calculator

1. In the calculator's set-up, switch the frequency column on (it is often called "Frequency" or "Freq"). Choose statistics mode for one variable, and clear any old data.

2. Key in each value xx (for grouped data, each mid-value) and its frequency ff.

3. Open the results. Check that nn equals Σf\Sigma f; then read xˉ\bar{x} and σx (xσn), as in lesson 3.5. Not sx.

With the frequency column on, the calculator's Σx\Sigma x is Σfx\Sigma fx and its Σx2\Sigma x^2 is Σfx2\Sigma fx^2.

With the frequency column off, the calculator takes each value once, and the answer is wrong.

Example 6: The 80 basketball players

The masses, mm kg, of the 80 basketball players of lessons 3.1 and 3.2 are grouped in the table. Use a calculator to estimate the mean and the standard deviation of their masses.

Mass, mm kgPlayers
70≤m<7570 \le m < 7533
75≤m<8075 \le m < 8088
80≤m<8580 \le m < 851616
85≤m<9085 \le m < 902121
90≤m<9590 \le m < 951919
95≤m<10095 \le m < 1001313
Total8080

Solution

The mid-values are 72.5, 77.5, 82.5, 87.5, 92.5 and 97.5.

Key in the six mid-values with the frequencies 3, 8, 16, 21, 19 and 13. The calculator shows n=80n = 80 (it agrees with Σf\Sigma f), Σx=7020\Sigma x = 7020, Σx2=619 650\Sigma x^2 = 619\,650, xˉ=87.75\bar{x} = 87.75 and σx =6.75= 6.75.

Mass, mm kgffxxfxfx
70≤m<7570 \le m < 753372.572.572.5×3=217.572.5 \times 3 = 217.5
75≤m<8075 \le m < 808877.577.577.5×8=62077.5 \times 8 = 620
80≤m<8580 \le m < 85161682.582.582.5×16=132082.5 \times 16 = 1320
85≤m<9085 \le m < 90212187.587.587.5×21=1837.587.5 \times 21 = 1837.5
90≤m<9590 \le m < 95191992.592.592.5×19=1757.592.5 \times 19 = 1757.5
95≤m<10095 \le m < 100131397.597.597.5×13=1267.597.5 \times 13 = 1267.5
Total808070207020

The table checks the calculator's Σx\Sigma x, which is Σfx\Sigma fx. By the second form: σ≈619 65080−87.752=7745.625−7700.0625=45.5625=6.75.\sigma \approx \sqrt{\dfrac{619\,650}{80} - 87.75^2} = \sqrt{7745.625 - 7700.0625} = \sqrt{45.5625} = 6.75\text{.}

The estimated mean, 87.75 kg, is close to the median read from the curve in lesson 3.2, about 88 kg.

The mean mass is about 87.75 kg and the standard deviation about 6.75 kg (estimates).

Common Mistakes

  • Dividing by the number of rows (or classes) in place of Σf\Sigma f, the number of values. In Example 4 the divisor is 20 patients, not 5 classes.
  • Working out (fx)2(fx)^2 in place of fx2=f×x2fx^2 = f \times x^2. For x=3x = 3, f=10f = 10 in Example 2: fx2=90fx^2 = 90, but (fx)2=900(fx)^2 = 900.
  • Using a class boundary (the lower or upper end of a class) in place of the mid-value.
  • Keying the values into the calculator without the frequency column. For Example 1 that gives the standard deviation of 0, 1, 2, 3, 4, each once: 1.41, not 1.16.
  • Rounding the mean before the second form. Keep 87.75 exact, not 88, and keep a recurring mean such as 27.166… in the calculator's memory.
  • Giving the standard deviation of grouped data as exact. It is an estimate, as the mean is.
  • Taking sx (xσn−1) from the calculator. Take σx (xσn).

Practice Questions

Try each question first, then open the solution to check your working. Give each answer to 3 significant figures unless it is exact, and say when it is an estimate.

Question 1

The table shows the number of pets, xx, owned by each of 30 pupils. Find the mean and the standard deviation by the first form of the formula.

Number of pets, xxPupils
001010
111212
2266
3322
Total3030
ShowHide solution

Σfx=0+12+12+6=30,\Sigma fx = 0 + 12 + 12 + 6 = 30\text{,} so xˉ=3030=1\bar{x} = \dfrac{30}{30} = 1, and the deviations are −1-1, 0, 1 and 2.

xxffx−xˉx - \bar{x}f(x−xˉ)2f(x - \bar{x})^2
001010−1-110×1=1010 \times 1 = 10
1112120012×0=012 \times 0 = 0
2266116×1=66 \times 1 = 6
3322222×4=82 \times 4 = 8
Total30302424

Σf(x−xˉ)2=10+0+6+8=24.\Sigma f(x - \bar{x})^2 = 10 + 0 + 6 + 8 = 24\text{.}

σ=2430=0.8=0.894…\sigma = \sqrt{\dfrac{24}{30}} = \sqrt{0.8} = 0.894\ldots

Mean 1 pet; standard deviation about 0.894 pets.

Question 2

Thirty pupils answered a quiz marked out of 5. The table shows their marks. Find the mean and the standard deviation of the marks by the second form of the formula.

Mark, xxPupils
1133
2266
331010
4477
5544
Total3030
ShowHide solution
xxfffxfxfx2fx^2
11331×3=31 \times 3 = 33×12=33 \times 1^2 = 3
22662×6=122 \times 6 = 126×22=246 \times 2^2 = 24
3310103×10=303 \times 10 = 3010×32=9010 \times 3^2 = 90
44774×7=284 \times 7 = 287×42=1127 \times 4^2 = 112
55445×4=205 \times 4 = 204×52=1004 \times 5^2 = 100
Total30309393329329

xˉ=9330=3.1\bar{x} = \dfrac{93}{30} = 3.1

σ=32930−3.12=10.966…−9.61=1.356…=1.164…\sigma = \sqrt{\dfrac{329}{30} - 3.1^2} = \sqrt{10.966\ldots - 9.61} = \sqrt{1.356\ldots} = 1.164\ldots

Mean 3.1 marks; standard deviation about 1.16 marks.

Question 3

The heights, hh cm, of 40 seedlings in a school garden are grouped in the table. Estimate the mean and the standard deviation of their heights.

Height, hh cmSeedlings
10≤h<2010 \le h < 2044
20≤h<3020 \le h < 3099
30≤h<4030 \le h < 401414
40≤h<5040 \le h < 5088
50≤h<6050 \le h < 6055
Total4040
ShowHide solution

Mid-values 15, 25, 35, 45, 55; then a frequency table of the mid-values.

Height, hh cmffxx
10≤h<2010 \le h < 20441515
20≤h<3020 \le h < 30992525
30≤h<4030 \le h < 4014143535
40≤h<5040 \le h < 50884545
50≤h<6050 \le h < 60555555
Total4040
xxfffxfxfx2fx^2
15154415×4=6015 \times 4 = 604×152=9004 \times 15^2 = 900
25259925×9=22525 \times 9 = 2259×252=56259 \times 25^2 = 5625
3535141435×14=49035 \times 14 = 49014×352=1715014 \times 35^2 = 17150
45458845×8=36045 \times 8 = 3608×452=162008 \times 45^2 = 16200
55555555×5=27555 \times 5 = 2755×552=151255 \times 55^2 = 15125
Total4040141014105500055000

xˉ≈141040=35.25\bar{x} \approx \dfrac{1410}{40} = 35.25 cm

σ≈55 00040−35.252=1375−1242.5625=132.4375=11.50…\sigma \approx \sqrt{\dfrac{55\,000}{40} - 35.25^2} = \sqrt{1375 - 1242.5625} = \sqrt{132.4375} = 11.50\ldots cm

Mean about 35.25 cm; standard deviation about 11.5 cm (estimates).

Question 4

Nurul keyed the values 0, 1, 2, 3, 4 of Example 1 into her calculator and read σx =1.41= 1.41. Explain why her answer is wrong.

ShowHide solution

She did not switch the frequency column on (or left every frequency as 1), so the calculator took each number of children once: five values in place of 40. Its nn shows 5, not 40.

The standard deviation of 0, 1, 2, 3, 4 is 4+1+0+1+45=2=1.41.\sqrt{\dfrac{4 + 1 + 0 + 1 + 4}{5}} = \sqrt{2} = 1.41\text{.} The 40 families give 1.35≈1.16\sqrt{1.35} \approx 1.16.

The frequencies were left out: each value counted once. With them, σ≈1.16\sigma \approx 1.16.

Question 5

For the times, xx minutes, that 50 pupils spent on homework one evening, Σf=50\Sigma f = 50, Σfx=1200\Sigma fx = 1200 and Σfx2=30 450\Sigma fx^2 = 30\,450. Find the mean and the standard deviation.

ShowHide solution

xˉ=120050=24\bar{x} = \dfrac{1200}{50} = 24 minutes.

σ=30 45050−242=609−576=33=5.744…\sigma = \sqrt{\dfrac{30\,450}{50} - 24^2} = \sqrt{609 - 576} = \sqrt{33} = 5.744\ldots minutes.

Mean 24 minutes; standard deviation about 5.74 minutes.

Question 6

The table shows the number of books, xx, that some pupils borrowed in a week. The mean is 2.5 books. Find kk, and then the standard deviation.

Number of books, xxPupils
1133
22kk
3366
4422
ShowHide solution

Σf=3+k+6+2=11+k\Sigma f = 3 + k + 6 + 2 = 11 + k and Σfx=3+2k+18+8=29+2k.\Sigma fx = 3 + 2k + 18 + 8 = 29 + 2k\text{.}

29+2k11+k=2.5,\dfrac{29 + 2k}{11 + k} = 2.5\text{,} so 29+2k=27.5+2.5k,29 + 2k = 27.5 + 2.5k\text{,} 1.5=0.5k1.5 = 0.5k and k=3k = 3.

xxfffxfxfx2fx^2
11331×3=31 \times 3 = 33×12=33 \times 1^2 = 3
22332×3=62 \times 3 = 63×22=123 \times 2^2 = 12
33663×6=183 \times 6 = 186×32=546 \times 3^2 = 54
44224×2=84 \times 2 = 82×42=322 \times 4^2 = 32
Total14143535101101

σ=10114−2.52=7.214…−6.25=0.964…=0.981…\sigma = \sqrt{\dfrac{101}{14} - 2.5^2} = \sqrt{7.214\ldots - 6.25} = \sqrt{0.964\ldots} = 0.981\ldots

k=3k = 3; σ≈0.982\sigma \approx 0.982 books.

Question 7

The table shows the times, tt minutes, that 60 commuters spent on the MRT on their way to work. (a) Estimate the mean and the standard deviation of the times. (b) Why are your answers estimates?

Time, tt minutesCommuters
0≤t<100 \le t < 1055
10≤t<2010 \le t < 201212
20≤t<3020 \le t < 302020
30≤t<4030 \le t < 401414
40≤t<5040 \le t < 5066
50≤t<6050 \le t < 6033
Total6060
ShowHide solution

(a) Mid-values 5, 15, 25, 35, 45, 55. By calculator with the frequency column: n=60n = 60, Σx=1630\Sigma x = 1630, Σx2=53 700\Sigma x^2 = 53\,700.

Time, tt minffxx
0≤t<100 \le t < 105555
10≤t<2010 \le t < 2012121515
20≤t<3020 \le t < 3020202525
30≤t<4030 \le t < 4014143535
40≤t<5040 \le t < 50664545
50≤t<6050 \le t < 60335555
Total6060
xxfffxfxfx2fx^2
55555×5=255 \times 5 = 255×52=1255 \times 5^2 = 125
1515121215×12=18015 \times 12 = 18012×152=270012 \times 15^2 = 2700
2525202025×20=50025 \times 20 = 50020×252=1250020 \times 25^2 = 12500
3535141435×14=49035 \times 14 = 49014×352=1715014 \times 35^2 = 17150
45456645×6=27045 \times 6 = 2706×452=121506 \times 45^2 = 12150
55553355×3=16555 \times 3 = 1653×552=90753 \times 55^2 = 9075
Total6060163016305370053700

xˉ≈163060=27.166…\bar{x} \approx \dfrac{1630}{60} = 27.166\ldots minutes.

σ≈53 70060−(163060)2=895−738.02…=156.97…=12.52…\sigma \approx \sqrt{\dfrac{53\,700}{60} - \left(\dfrac{1630}{60}\right)^2} = \sqrt{895 - 738.02\ldots} = \sqrt{156.97\ldots} = 12.52\ldots minutes. (Keep the mean in the calculator's memory; do not round it to 27.2 first.)

(b) Only the classes are known, not the single times. Each time was taken to be at the middle of its class.

(a) Mean about 27.2 minutes; standard deviation about 12.5 minutes (b) because each time was replaced by the mid-value of its class.

Question 8

Wei Jie estimated the standard deviation of the waiting times of Example 4 like this: "13 1005−23.52".\sqrt{\dfrac{13\,100}{5} - 23.5^2}\text{".} What did he do wrong, and how could he have seen that his answer was wrong?

ShowHide solution

He divided Σfx2\Sigma fx^2 by 5, the number of classes, in place of Σf=20\Sigma f = 20, the number of patients.

His working gives 2620−552.25=2067.75≈45.5\sqrt{2620 - 552.25} = \sqrt{2067.75} \approx 45.5 minutes: more than the largest distance of any mid-value from the mean (21.5), which a standard deviation can never be (lesson 3.5).

He divided by the number of classes, 5, not by Σf=20\Sigma f = 20. The right estimate is about 10.1 minutes.

Question 9

Every frequency in the table of Example 1 is doubled (80 families, in the same proportions). Without working, say what happens to the mean and the standard deviation. Then check with the second form.

ShowHide solution

The mean and the standard deviation depend only on the values and on what fraction of the data each value is. Doubling every frequency keeps the fractions, so neither changes.

Check: Σf=80\Sigma f = 80, Σfx=160\Sigma fx = 160, Σfx2=428\Sigma fx^2 = 428. xˉ=16080=2\bar{x} = \dfrac{160}{80} = 2 and σ=42880−4=1.35,\sigma = \sqrt{\dfrac{428}{80} - 4} = \sqrt{1.35}\text{,} as before.

Neither changes: the mean is still 2 and σ\sigma still about 1.16.

Question 10

A teacher has two records of the same 20 runners' times: the list of 20 times, and the grouped table of Example 5. Which should she use to give the standard deviation, and why? When must she use the table?

ShowHide solution

The list: it gives the true standard deviation, 5.05 s, while the table gives only an estimate, 5.58 s, because it replaces each time by a mid-value.

She must use the table (and say that the answer is an estimate) when only the grouped data are kept, for example when the single times were never written down.

The list, which gives the exact value; the table only when the single values are not known, and then the answer is an estimate.

Key Points

  • For a frequency table: xˉ=ΣfxΣf\bar{x} = \dfrac{\Sigma fx}{\Sigma f} and σ=Σf(x−xˉ)2Σf=Σfx2Σf−xˉ2.\sigma = \sqrt{\dfrac{\Sigma f(x - \bar{x})^2}{\Sigma f}} = \sqrt{\dfrac{\Sigma fx^2}{\Sigma f} - \bar{x}^2}\text{.}
  • fx2fx^2 is f×x2f \times x^2, not (fx)2(fx)^2; divide by Σf\Sigma f, not by the number of rows.
  • For grouped data, xx is the mid-value of each class, and the mean and standard deviation are estimates.
  • On a calculator, switch the frequency column on, key in each value (or mid-value) with its frequency, check n=Σfn = \Sigma f, and take σx.
  • Keep the exact mean until the last line; give σ\sigma to 3 significant figures with its units.

Previous lesson3.5 Standard Deviation Next lesson3.7 Comparing Two Sets of Data

All Secondary 4 (Grade 10) lessons

Home Campus has lessons for Primary 3 to Secondary 4, and practice questions with worked solutions and worksheets for Primary 3 to Secondary 2. Register free to practise and keep track of progress.

This lesson follows the Singapore Secondary 4 syllabus and was checked question by question before it was published. Spotted a mistake? Tell us.