AP Statistics Topic 2.9: Parameters of Random Variables

01 · Start here

The concept in 60 seconds

A random variable can give different values on different trials. A parameter describes a fixed characteristic of its probability distribution.

Two useful parameters are the mean, which describes the long-run average outcome, and the standard deviation, which describes the typical size of deviations from that mean.

Visual guide 1: random outcomes, fixed parameters

1

session $X$ can change

The next session may produce $0, 1, 2,\text{ or }3$ completed problems.

That single result is an outcome.

Long-run average $\mu_X = 1.9$

The model has a fixed mean of $1.9$ problems per session.

This need not be a possible single count.

Spread around the mean $\sigma_X \approx 0.943$

The count typically differs from $1.9$ by about $0.943$ problems.

This is also fixed for the stated model.

Original fictional study-session model, developed below. Parameters describe its distribution; finite sample averages can vary.

For a given model, the parameters stay fixed even when the next outcome is uncertain. A sample average computed from a finite set of trials can vary; it is not automatically equal to the model’s mean.

Think of two separate questions:
“What happens in this $1$ trial?” asks for a random outcome.
“What average and spread does the model describe?” asks for parameters.

Quick check: must expected value be a possible outcome?

No. A mean summarizes a distribution; it does not have to be one of its possible values. A count variable can have mean $1.9$ even though an individual count cannot be $1.9$ .

02 · One clear variable and context

The study-session model

Consider an original fictional model for students’ short study sessions. Let $X$ be the number of practice problems completed in $1$ randomly selected session. The model assigns these probabilities:

Original fictional model: $X$ is the number of practice problems completed in $1\text{ randomly selected short study session}$ .
Completed problems $x$	Probability $P(X = x)$
$0$	$0.10$
$1$	$0.20$
$2$	$0.40$
$3$	$0.30$

The possible values are $\{0, 1, 2, 3\}$ . All probabilities are between $0$ and $1$ , and $0.10 + 0.20 + 0.40 + 0.30 = 1$ , so this is a valid distribution.

$1$ trial is $1$ complete study session, not $1$ problem. The units of $X$ are problems completed per session.

The value $2$ is most likely because it has the largest probability, $0.40$ . That makes it the mode. The mean answers a different question and uses every possible value and its probability.

03 · Separate center, spread and units

Key ideas and notation

Mean / expected value

Written $\mu_X$ or $\operatorname{E}(X)$ . It is the probability-weighted average of the possible values.

For the study model: $\mu_X = 1.9$ problems per session.

Variance

Written $\sigma_X^2$ or $V(X)$ . It averages squared deviations from the mean, using probability weights.

For the study model: $0.89$ in squared problem-count units.

Standard deviation

Written $\sigma_X$ or $\operatorname{SD}(X)$ . It is the square root of the variance.

For the study model: about $0.943$ problems per session.

Parameter versus statistic

A parameter describes the distribution or population. A statistic is calculated from sample observations.

Model mean $\mu_X$ is fixed; sample mean $\bar{x}$ can change across samples.

The symbol $\sum$ means “add over all possible values.” Each value $x$ is paired with its own probability $P(X = x)$ .

Mean:

\mu_X=\sum_x xP(X=x)

Variance:

\sigma_X^2=\sum_x(x-\mu_X)^2P(X=x)

Standard deviation:

\sigma_X=\sqrt{\sum_x(x-\mu_X)^2P(X=x)}

The mean and SD have the same units as $X$ . Variance has squared units. SD cannot be negative; it is $0$ when the variable is constant with probability $1$ .

04 · Multiply each value by its probability

Calculate expected value

Check the distribution and identify the units.
Multiply each possible value $x$ by its probability.
Add all the products.
Interpret the result as a long-run average in context.

\begin{aligned}\mu_X&=0(0.10)+1(0.20)+2(0.40)+3(0.30)\\&=0+0.20+0.80+0.90\\&=1.90\text{ problems per session}.\end{aligned}

Visual guide 2: probabilities weight the average

Imagine $100$ sessions with exactly this composition: it makes the probability weights easier to see.

$10$ sessions complete $0$ problems · $10\%$

$20$ sessions complete $1$ problem · $20\%$

$40$ sessions complete $2$ problems · $40\%$

$30$ sessions complete $3$ problems · $30\%$

Total problems: $10 \times (0) + 20 \times (1) + 40 \times (2) + 30 \times (3) = 190$ .
Average: $190 ÷ 100 = 1.9$ problems per session.

The same weighted average calculated directly from the model.
Value $x$	Probability $p(x)$	Mean contribution $x \cdot p(x)$
$0$	$0.10$	$0.00$
$1$	$0.20$	$0.20$
$2$	$0.40$	$0.80$
$3$	$0.30$	$0.90$
Total	$1.00$	$1.90$

The $100$ -session composition is an illustration, not observed data or a prediction of exact future frequencies. Bar lengths use a common $0\%\text{–}100\%$ scale. Expected value comes from the probability weights, regardless of a finite run’s exact counts.

Why a plain average of the listed values is wrong here

$\frac{0 + 1 + 2 + 3}{4} = 1.5$ gives each value an equal weight of $\frac{1}{4}$ . The model gives them different probabilities, so that is not its mean.

An unweighted average works for a finite list of equally likely values. In general, use the probabilities supplied by the model.

Visual guide 3: the mean lies between possible values

Exact fictional model. The dashed mean line marks the weighted average; it is not an extra possible outcome or an extra probability. The tallest bar is at $2$ , while the mean is $1.9$ .

The dashed reference marks $\mu_X = 1.9$ . There is no bar at $1.9$ because it is not a possible count. The mean is also different from the mode, which is $2$ .

Reasonableness check: for this finite distribution, the mean must be between its minimum $0$ and maximum $3$ . The answer $1.9$ passes that check; an answer of $4$ would signal an error.

05 · Use weighted squared deviations

Calculate variance and standard deviation

$2$ distributions can have the same mean but different spreads. To measure spread, examine how far the possible values lie from their mean and how likely those values are.

Start with the unrounded mean. Here $\mu_X = 1.9$ exactly.
Subtract the mean from each value. This gives the signed deviation $x – \mu_X$ .
Square each deviation. Squaring prevents negative and positive deviations from cancelling.
Multiply each square by its probability. Frequent outcomes contribute more weight.
Add the weighted squares. The sum is variance.
Take $1\text{ square root}$ at the end. The result is standard deviation.

Visual guide 4: weight each squared deviation

Variance calculation for the study-session model, using $\mu_X = 1.9$ .
Value $x$	Probability $p(x)$	Deviation $x – 1.9$	Squared deviation	Weighted squared deviation
$0$	$0.10$	$-1.9$	$3.61$	$0.361$
$1$	$0.20$	$-0.9$	$0.81$	$0.162$
$2$	$0.40$	$+0.1$	$0.01$	$0.004$
$3$	$0.30$	$+1.1$	$1.21$	$0.363$
Total	$1.00$	—	—	$0.890$

First find the center $\text{Mean} = 1.9$

Use the probability-weighted mean before finding deviations.

Add the last column $\text{Variance} = 0.89$

Variance uses squared problem-count units.

Take

1\text{ square root}

$\text{SD} \approx 0.943$

Standard deviation returns to the original problem-count units.

Each deviation is squared before it is weighted. Add all $4$ weighted squared deviations, then take the square root of that sum. The signed deviations would cancel in their weighted sum; their squares do not.

The weighted squared deviations sum to $0.89$ . Therefore:

\sigma_X^2=0.89

in squared problem-count units.

\sigma_X=\sqrt{0.89}\approx0.943

problems per session.

Why not average the signed deviations?

For this model, $\sum_x(x-1.9)P(X=x)=-0.19-0.18+0.04+0.33=0$ . The values are spread out, but the signed deviations cancel. Variance avoids that cancellation by squaring first.

SD is based on squared deviations; it is not the probability-weighted average of absolute distances. “Typical deviation” is an interpretation of SD, not a replacement calculation.

No $n-1$ correction

We are calculating a parameter from the full stated probability distribution. Use its probabilities as weights. The sample-SD formula with $n – 1$ is for a different task involving sample data.

06 · Explain what the results say

Interpret the parameters in context

Mean: “Under this model, the long-run average number of practice problems completed is $1.9$ per study session.”

Standard deviation: “The number completed in a session typically differs from the mean of $1.9$ by about $0.943$ problems.”

An individual session still has $0, 1, 2,\text{ or }3$ completed problems. The mean does not predict exactly what the next session will produce.

A statement about SD describes the distribution’s spread. It does not say every session is exactly $0.943$ from the mean, that $94.3\%$ of sessions meet a condition, or that all values fall within $1$ SD.

Long-run average does not mean exact agreement after a fixed number of trials

The graph below shows $1\text{ actual computer-generated simulation}$ of this fictional model. Each complete simulated session selects uniformly from $10$ tickets carrying values $0, 1, 1, 2, 2, 2, 2, 3, 3, 3$ . These tickets reproduce the model’s probabilities, and the running mean uses all sessions up to that point.

Visual guide 5: one recorded run and the long-run mean

      EXCELLENT  
   Based on 51 reviews 
    
 
 
 
   
 
 
       Posted on Google  
   
   0000003998 : Abdad Alam Shamim Alam 
 
 
 
 Trustindex verifies that the original source of the review is Google. This is the bestest institution I have ever come to and I love it very much.
  
 
      Posted on Google  
   
   Aliki S 
 
 
 
 Trustindex verifies that the original source of the review is Google. 
  
 
      Posted on Google  
   
   Ali 
 
 
 
 Trustindex verifies that the original source of the review is Google. Sly is him
  
 
      Posted on Google  
   
   Jitendra Kumar Kumawat 
 
 
 
 Trustindex verifies that the original source of the review is Google. 
  
 
      Posted on Google  
   
   Raahil Hasan 
 
 
 
 Trustindex verifies that the original source of the review is Google. 
  
 
      Posted on Google  
   
   Tina Mudarres 
 
 
 
 Trustindex verifies that the original source of the review is Google. The classes are amazing and my child learnt so much
  
 
      Posted on Google  
   
   alisha gadoya 
 
 
 
 Trustindex verifies that the original source of the review is Google. I have experienced a lot of good things, it has taught me so many things and from B/Cs I have gone to an A! This is wonderful and Mavish’s class is awesome.
  
 
      Posted on Google  
   
   smasher 123 
 
 
 
 Trustindex verifies that the original source of the review is Google. It’s sensational my kid went. First he only would get C now it’s all A’s really good and recommend
  
 
      Posted on Google  
   
   Mosa Al- Samaraie 
 
 
 
 Trustindex verifies that the original source of the review is Google. AMAZING PRICES GREAT TEACHERS STRAIGHT FORWARD LEARNING STEADY PACE IN TUTORING
  
 
      Posted on Google  
   
   Cael Dagnelie 
 
 
 
 Trustindex verifies that the original source of the review is Google. I had a great experience learning math with Ms. Mavish. She explains complex topics in a very clear and simple way, which made it easier for me to understand and enjoy the subject. Her patience and dedication really stood out, and she always made sure that everyone in the class was keeping up. I especially appreciated how approachable she was — I never felt afraid to ask questions, and she was always willing to help. Thanks to her teaching, my confidence in math has grown a lot. I’m really thankful for the effort she puts into every lesson!

NUM8ERS is one of finest tutoring institutes in UAE, Located in Al Barsha 1, Dubai. Close to DUBAI AMERICAN ACADEMY (DAA) & AMERICAN SCHOOL OF DUBAI (ASD).

Parameters of Random Variables

By the end of this lesson, you should be able to:

The concept in 60 seconds

The study-session model