Common Errors in Expected Value and Variability

Small Calculation Errors Can Change the Meaning

A probability distribution describes both the possible values of a random variable and how likely each value is. A calculation can use the right values and still go wrong if it treats unlikely and likely outcomes equally, leaves deviations unsquared, or confuses standard deviation with variance. These mistakes do not just affect arithmetic: they can change what the result says about the distribution.

In The Mean of a Discrete Random Variable, we found the expected value by multiplying each possible value by its probability and adding. In Variance of a Discrete Random Variable and The Standard Deviation of a Discrete Random Variable, we measured spread using squared deviations from the mean, then took the square root to return to the original units. And in the previous tutorial, Standard Deviation of the Sum of Independent Variables, we combined independent variables by adding their variances—not their standard deviations.

This tutorial focuses on diagnosing those common errors. A useful habit is to check not only whether the arithmetic is correct, but also whether each calculation matches the meaning and units of the quantity you are finding.

Key reminder: Expected value weights values by probability. Variance weights squared deviations from the mean by probability. Standard deviation is the square root of variance. For independent random variables, variances—not standard deviations—add.

Error 1: Averaging the Possible Values Equally

The expected value is a probability-weighted average, not usually the ordinary average of the listed \(x\)-values. Each value’s contribution depends on its probability. Giving every value equal weight is appropriate only when all possible values are equally likely.

A quick diagnostic is to ask whether the calculation includes every probability. If it does not, it may be averaging the values as though each had the same chance, even when the distribution says otherwise. The value with the highest probability should generally have more influence on the expected value than a rare value.

Worked Example: Do Not Average Unequally Likely Values

A fictional game assigns a player a score \(X\) of 0, 2, or 8 points. The probabilities are 0.50, 0.30, and 0.20, respectively. Find the expected score and explain why the ordinary average of the three scores is not the answer.

State. \(X\) is the score, in points, from one play. Its distribution is valid because each probability is between 0 and 1 and \(0.50+0.30+0.20=1\). We want \(\mu_X\), the expected score.

Plan. Use the probability-weighted mean from The Mean of a Discrete Random Variable: multiply each possible score by its probability and add the products. Do not give the three scores equal weight unless their probabilities are equal.

Do.

\begin{aligned} \mu_X &=(0)(0.50)+(2)(0.30)+(8)(0.20)\\ &=0+0.60+1.60\\ &=2.20\text{ points}. \end{aligned}

As a check, the contributions are 0 points from the score of 0, 0.60 points from the score of 2, and 1.60 points from the score of 8; their sum is 2.20 points. The ordinary average of the listed values would be \((0+2+8)/3=10/3\approx3.3333\) points. That calculation gives each score equal weight and ignores the probabilities, so it does not find the expected score for this distribution.

Conclude. The expected score is 2.20 points per play. It is the probability-weighted center and, over many repetitions under the same model, the long-run average score—not a promise that one play will produce 2.20 points.

It is possible for an expected value not to be one of the variable’s possible values. That is not an error: it is a weighted center of the distribution. The relevant check is whether the probabilities and values were paired correctly and whether all the weighted contributions were included.

Error 2: Forgetting to Square Deviations

Variance is not the probability-weighted average of signed deviations \(x-\mu_X\). Positive and negative deviations would cancel, giving a weighted total of zero. Instead, variance uses \((x-\mu_X)^2\) for each value. Squaring makes every contribution nonnegative and gives larger deviations more influence than smaller ones.

A related mistake is to use absolute deviations, \(|x-\mu_X|\), as if they were variance. Absolute deviations do not cancel, but their weighted average is not the variance or standard deviation. Follow the calculation described in Variance of a Discrete Random Variable: subtract the mean, square the result, multiply by the probability, and add.

Worked Example: Keep the Squared Deviations

A fictional sensor records an error \(X\), in dollars, of 10, 20, or 30 dollars. The probabilities are 0.25, 0.50, and 0.25. Find the variance and standard deviation. Also identify the result of mistakenly using absolute deviations instead of squared deviations.

State. \(X\) is the sensor’s error, in dollars. The probabilities are valid because they are between 0 and 1 and sum to \(0.25+0.50+0.25=1\). We need \(\mu_X\), \(\sigma_X^2\), and \(\sigma_X\).

Plan. First find the expected error. Then calculate each deviation from that mean, square it, weight it by its probability, and add to find variance. Take the square root of the variance for standard deviation.

Do. The expected error is:

\begin{aligned} \mu_X &=(10)(0.25)+(20)(0.50)+(30)(0.25)\\ &=2.5+10+7.5\\ &=20\text{ dollars}. \end{aligned}

The mean is also centered symmetrically between the equally likely values 10 and 30, with the middle value 20 carrying probability 0.50. Now weight the squared deviations:

\begin{aligned} \sigma_X^2 &=(10-20)^2(0.25)+(20-20)^2(0.50)+(30-20)^2(0.25)\\ &=100(0.25)+0(0.50)+100(0.25)\\ &=25+0+25\\ &=50\text{ dollars}^2. \end{aligned}

As an arithmetic check, the two nonzero contributions are each \(100(0.25)=25\), so the variance is \(25+25=50\) square dollars. A second check uses the weighted squared values: \((10^2)(0.25)+(20^2)(0.50)+(30^2)(0.25)=25+200+225=450\). Subtracting the square of the mean, \(20^2=400\), gives \(450-400=50\), the same variance.

The standard deviation is:

\sigma_X=\sqrt{50}\approx7.0711\text{ dollars}.

The result checks because \(7.0711^2\) is approximately 50, allowing for rounding. If we mistakenly used absolute deviations, the weighted average would be \((10)(0.25)+(0)(0.50)+(10)(0.25)=5\) dollars. That value is an average absolute deviation, not the variance or standard deviation.

Conclude. The variance is 50 square dollars, and the standard deviation is approximately 7.0711 dollars. In context, errors typically differ from the mean error of 20 dollars by about 7.0711 dollars.

Notice the units. Squaring a deviation measured in dollars produces square dollars, so variance is in square dollars. Taking the square root returns to dollars, the original units of \(X\). Reporting a variance as though it were a typical distance, or reporting standard deviation in squared units, confuses two different summaries.

Error 3: Adding Standard Deviations

When two independent random variables are added, their means add, as explained in Mean of the Sum of Two Random Variables. Their standard deviations do not generally add. The rule from Standard Deviation of the Sum of Independent Variables is to add variances and then take the square root:

Formula: If \(X\) and \(Y\) are independent, then

\sigma_{X+Y}=\sqrt{\sigma_X^2+\sigma_Y^2}.

Adding \(\sigma_X+\sigma_Y\) is not the independent-sum rule. The independence condition matters.

A common source of confusion is that standard deviations use the same units as the variables, which can make direct addition look plausible. But the rule combines squared distances first. The variances have squared units, and the final square root returns the result to the original units.

Worked Example: Do Not Add Independent Standard Deviations

A fictional community event has two independent setup tasks. Let \(X\) be the first task’s time and \(Y\) the second task’s time, both measured in minutes. The model gives \(\mu_X=12\), \(\mu_Y=8\), \(\sigma_X=3\), and \(\sigma_Y=4\). Find the mean and standard deviation of the total setup time \(X+Y\).

State. \(X+Y\) is the combined setup time in minutes. The model states that the task times are independent, so the independent-sum standard-deviation rule applies.

Plan. Add the means to find the mean total. For spread, square each standard deviation, add the variances, and take the square root. Do not add the two standard deviations directly.

Do. The mean total is:

\mu_{X+Y}=\mu_X+\mu_Y=12+8=20\text{ minutes}.

The two mean contributions add to 20 minutes, which can be checked by \(12+8=20\). For the standard deviation, first find and combine the variances:

\begin{aligned} \sigma_{X+Y} &=\sqrt{\sigma_X^2+\sigma_Y^2}\\ &=\sqrt{3^2+4^2}\\ &=\sqrt{9+16}\\ &=\sqrt{25}\\ &=5\text{ minutes}. \end{aligned}

The variance calculation checks because \(9+16=25\) square minutes and \(5^2=25\). The tempting but incorrect sum \(3+4=7\) minutes does not follow the rule for independent variables.

Conclude. Under this model, the total setup time has mean 20 minutes and standard deviation 5 minutes. Total setup times typically differ from their mean by about 5 minutes. This spread calculation depends on the stated independence of the task times.

If variables are not independent, do not apply the independent-sum formula automatically. As emphasized in the previous tutorial, dependence can affect the spread of a total. Check the model or context before combining standard deviations.

A Practical Error-Checking Routine

A short audit can catch many mistakes before you report an answer. It does not replace showing the calculation; it helps confirm that the calculation answers the question being asked.

Check the probabilities.
Confirm that the distribution includes the possible values and that the probabilities are between 0 and 1 and sum to 1.

Match the operation to the summary.
For expected value, multiply each value by its probability. For variance, use squared deviations from the mean and multiply each by its probability.

Check the units.
Expected value and standard deviation use the original units. Variance uses squared units.

Check how variables are combined.
Means of a sum add. For independent variables, add variances to find the variance of the sum, then take the square root.

Magnitude can also be a useful warning signal. A probability-weighted mean should lie between the smallest and largest possible values. A standard deviation cannot be negative. And after calculating a variance, its square root should square back to the variance, allowing for rounding. These checks do not prove that every step is right, but they can reveal an answer that is incompatible with the distribution or its units.

Common Mistakes and AP Exam Tips

Taking the ordinary average of the possible values. This gives equal weight to every listed value, whether or not the probabilities are equal. A full-credit expected-value calculation shows each product \(xP(X=x)\) and adds the products.
Leaving deviations signed. Positive and negative deviations can cancel. A full-credit variance calculation squares each difference before multiplying by its probability.
Using absolute deviations as variance. Absolute deviations avoid cancellation, but they are not squared deviations. Show the squared-deviation contributions required for variance.
Calling variance a standard deviation. Variance is in squared units; standard deviation is its nonnegative square root and uses the original units. Label both clearly if the question asks for both.
Adding standard deviations for independent variables. Add variances and take the square root. State that the variables are independent when using this rule.
Giving only a number. A strong response identifies the random variable, reports the appropriate units, and interprets the result in context. For standard deviation, explain it as a typical distance from the mean, not a guaranteed maximum distance.

On an AP response, make the route to the answer visible. For expected value, show the probability weights. For variability, show the deviations and their squares, followed by the weighted sum and square root where needed. A correct answer with unexplained arithmetic may not make clear whether you used the correct statistical method.

Key Takeaway

Most errors in these calculations come from treating probabilities as if they were irrelevant or from mixing up variance and standard deviation. Keep the role of each quantity clear: probabilities weight the outcomes, squared deviations measure spread for variance, and standard deviation returns that spread to the original units.

Key takeaway: Do not average possible values equally unless they are equally likely. Find variance using weighted squared deviations, and for independent sums add variances—not standard deviations—before taking the square root.

Check Your Understanding

For each question, show enough work to identify the method and explain any likely error.

A random variable takes values 1 and 9 with probabilities 0.75 and 0.25. Explain why the ordinary average of 1 and 9 is not its expected value, then find the expected value.
Why would adding the weighted signed deviations from a distribution’s mean fail to give its variance?
A variable has mean 5 units and takes values 3 and 7, each with probability 0.50. Find its variance and standard deviation, and state the units for each.
Two independent task times have standard deviations of 6 minutes and 8 minutes. Find the standard deviation of their total and explain why 14 minutes is not the answer.
For a discrete distribution, name one check for a calculated expected value and one check for a calculated standard deviation.