In: Statistics and Probability
Grade:ABCDF
Probability:0.10.30.40.10.1
To calculate student grade point averages, grades are expressed in a numerical scale with A = 4, B = 3, and so on down to F = 0.
Find the expected value. This is the average grade in this course.
Explain how to simulate choosing students at random and recording their grades. Simulate 50 students and find the mean of their 50 grades. Compare this estimate of the expected value with the exact expected value from part (a). (The law of large numbers says that the estimate will be very accurate if we simulate a very large number of students.)
Calculation for expected value-
Suppose, random variable X denotes numerical value corresponding to grade obtained.
So, we have probability mass function as follows.
Grade obtained |
A | B | C | D | F | Total |
Numerical scale value, |
4 | 3 | 2 | 1 | 0 | - |
Probability, |
0.10 | 0.30 | 0.40 | 0.10 | 0.10 | 1 |
So, expected value
= 0*0.10 + 1*0.10 +2*0.40 + 3*0.30 +4*0.10 = 2.2
Simulation-
We can perform simulation either by hand using a random number table or by using any statistical software.
Let us simulate by hand using random number table.
X | P(X=x) | Cumulative probability | Random number interval |
0 | 0.10 | 0.10 | 00-09 |
1 | 0.10 | 0.20 | 10-19 |
2 | 0.40 | 0.60 | 20-59 |
3 | 0.30 | 0.90 | 60-89 |
4 | 0.10 | 1 | 90-99 |
Note- As we are going to take last two digits we have taken interval 00-99 (instead of 01-100).
We now take last two digits of each number of the following random number table to use. One may use any other random number table or even this one considering any other way of choosing (e.g. first two digits of each random number).
Simulated numerical scale values are as follows.
Serial number | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
Random number | 64 | 18 | 61 | 74 | 20 | 15 | 91 | 31 | 78 | 54 |
Numerical scale value | 3 | 1 | 3 | 3 | 2 | 1 | 4 | 2 | 3 | 2 |
Serial number | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 |
Random number | 38 | 82 | 58 | 49 | 82 | 92 | 36 | 66 | 83 | 81 |
Numerical scale value | 2 | 3 | 2 | 2 | 3 | 4 | 2 | 3 | 3 | 3 |
Serial number | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 |
Random number | 44 | 73 | 32 | 14 | 84 | 63 | 72 | 63 | 04 | 83 |
Numerical scale value | 2 | 3 | 2 | 1 | 3 | 3 | 3 | 3 | 0 | 3 |
Serial number | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 |
Random number | 08 | 37 | 01 | 17 | 42 | 86 | 75 | 05 | 41 | 49 |
Numerical scale value | 0 | 2 | 0 | 1 | 2 | 3 | 3 | 0 | 2 | 2 |
Serial number | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 |
Random number | 56 | 60 | 16 | 71 | 78 | 25 | 17 | 15 | 07 | 31 |
Numerical scale value | 2 | 3 | 1 | 3 | 3 | 2 | 1 | 1 | 0 | 2 |
So, calculating number of occurrence of numerical scale values we get,
Numerical scale value | 0 | 1 | 2 | 3 | 4 | Total |
Frequency | 5 | 7 | 16 | 20 | 2 | 50 |
So, simulated mean = (0*5 + 1*7 +2*16 + 3*20 + 4*2) / 50 = 107 / 50 = 2.14
Conclusion-
We see expected mean (2.2) and simulated mean (2.14) are very close.
Note: If we simulate using another random number table or choosing number from this table with different procedure if choosing or using any statistical software, we will get different simulated mean values each time (even if we use same statistical software). But all these simulated values will be close to expected value.