3.1
Modeling Linear Relationships with Algebra
Learning Objectives
Upon completion of this section, you should be able to
- Identify elements in a linear model of the form
- Create a linear model with algebra between two quantitative variables
- Graph a linear model
- Solve application problems using a linear model created with algebra
Linear Models
Two variables are in linear relationship if we can describe how they change together by adding or subtracting a fixed constant for each unit increase in one variable. For example the cost for hiring a plumber may start with a $75 fee for coming to your house plus an hourly rate of $60. The two variable here are the total cost for the job and the number of hours needed to complete it. For each one hour increase to complete the job we would see an increase of $60 to the total cost.
What you will work on in this section is how to write an algebraic model to describe that relationship between two variables in applications. We will start with the general equation of a line in slope-intercept form.
Linear Model
A linear model (equation) has the form .
- b is the of the graph and indicates the point at which the graph crosses the .
- m is the slope of the line (also called the rate of change) and indicates the vertical displacement (rise) and horizontal displacement (run) between each successive pair of points. The formula for slope between two points is:
We call this form of a linear equation the slope-intercept form as it gives the information about the slope, m, and y-intercept, b.
Example 1
Which of the following represent a linear model. If it is a linear model identify the slope m and intercept b.
Solutions
- This is a linear model. Slope is 10 and the y-intercept is 20.
- This is not a linear model as we are squaring the x term in the equation.
- This is not a linear as the second term in the equation has x in the exponent.
- This is a linear model. We can rewrite it as . The slope is 2 and the y-intercept is 5.
- This is a linear model. We can rewrite it as . The slope is -45 and has a y-intercept of 100.
Before we look at applications lets review how to create a linear model for given sets of points and identify the key elements of slope and intercept.
Example 2
Find a linear model that goes through the points
Solution
When given two points you start with the slope formula to find m:
Now we were not given the y-intercept directly, so we will first update the linear model with our information about the slope:
Now to find b we can use either of the two given points to put into the equation and solve. We will use in the work below.
Putting our value in for b we get the linear model is:
Try it Now 1
Find the equation of a line that goes through the points .
Hint 1
Hint 2
Answer
Graph a Linear Model
We will start from a linear model and graph the line with two different strategies (plot by points and plot using information about the line from the given information of slope and intercept).
Graphing a linear model by plotting points
To find points on a graph for any model, we can choose input values and evaluate the model at these input values to find the corresponding output values. The input values and corresponding output values form coordinate pairs. We then plot the coordinate pairs on a grid.
In general, we should evaluate a linear model at a minimum of two inputs in order to find at least two points on the graph. For example, given the model, , we might use the input values 1 and 2. Evaluating the model for an input value of yields an output value of , which is represented by the point (1, 2). Evaluating the model for an input value of yields an output value of , which is represented by the point (2, 4). Choosing three points is often advisable because if all three points do not fall on the same line, we know we made an error.
Plot by Points
- Choose a minimum of two input values.
- Evaluate the model at each input value.
- Use the resulting output values to identify coordinate pairs.
- Plot the coordinate pairs on a grid.
- Draw a line through the points.
Example 3
Graph the linear model defined by:
Solution
Begin by choosing input values. This function includes a fraction with a denominator of 3, so let’s choose multiples of 3 as input values. We will choose 0, 3, and 6.
Evaluate the function at each input value, and use the output value to identify coordinate pairs.
Plot the coordinate pairs and draw a line through the points. The graph below is for the model
As expected we the graph is a line for the model. In addition, the graph has a downward slant when reading left to right over the horizontal axis, which indicates a negative slope.
Try it Now 2
Graph the linear model given by
Hint 1
Answer
Graphing a linear model by using slope and a point
Another way to graph linear models is to use specific characteristics of the linear model rather than plotting points. The first characteristic is the y-intercept, which is the point at which the input value is zero. To find the y-intercept, we can set in the equation.
The other characteristic of the linear model is the slope m, which is a measure of its steepness. Recall that the slope is the rate of change of output to input. The slope of a function is equal to the ratio of the change in outputs to the change in inputs. Another way to think about the slope is by dividing the vertical difference, or rise, by the horizontal difference, or run.
If we know the slope and either a point or the y-intercept it is possible to graph the linear model by placing the point on the coordinate system and using the property of the slope to find another point the graph goes through. After identifying that next point we can draw the line between the two points or continue to find points on the line using the slope. Choosing at least two points from the slope is often advisable because if all three points do not fall on the same line, we know we made an error.
Graphing a linear model from slope and a point
- If not given a point evaluate the linear model at to get the y-intercept.
- Identify the slope from the linear model or given information.
- Plot the point represented by the y-intercept.
- Use to determine at least one more points on the line (two or more is preferred).
- Sketch the line that passes through the points.
Example 4
Graph the following using information about the slope and the y-intercept:
Solution
The slope is . Because the slope is positive, we know the graph will slant upward from left to right. The y-intercept is the point on the graph when . The graph crosses the y-axis at . Now we know the slope and the y-intercept.
We can begin graphing by plotting the point . We also know that the slope is rise over run, . We know the slope for our linear model is , which means that the rise is 1 and the run is 2. So starting from our y-intercept , we can rise 1 and then run 2, or run 2 and then rise 1. We repeat until we have a few points, and then we draw a line through the points as shown below.
Do all linear models have a y-intercept? What about x-intercept?
The general answer to this is no. Some linear models could be in the form of or , where c is some constant. The first linear equation when graphed forms a vertical line which the second is a horizontal line. Both of these, however, is not of interest to us as we are looking at the relationship between two variables that are related. For our work in this section we will only look at linear equations that do not represent a vertical or horizontal line.
Try it Now 3
Graph the linear model by using information about the slope and y-intercept:
Hint 1
Hint 2
Answer
Solve applications problems involving a linear model
We will keep our focus in this area related to our study of statistics in that we will be looking at how to construct a model from two given data points from a sample or a population of interest. In addition we will place emphasis on understanding the model in terms of the application presented.
When starting our work for finding a linear model between two quantitative variables in applications we will have to decide which variable is treated as x in the linear equation and which ones as y. If it is known that one variable is more dependent on the other variable we will assign that dependent variable as y and the other will be called the independent (or explanatory) variable and assigned to x. We will often, however, choose constants beyond x and y if those constants help keep track of what each variable represents in the application.
For instance if we are looking at the relationship between the age of a child and the number of words in the vocabulary we may want to use "a" for age and "v" for the number of words in the vocabulary. This use of "a" and "v" makes it easier to recognize how the model works instead of trying to figure out what the letters "x" and "y" represent.
Find a linear model in applications
- Identify the variables in the given information.
- Assign the variables letter names.
- Identify if one variable is independent (or explanatory). If it is that variable is treated like x in our linear model. If there is no direct independent or dependent variable you can randomly assign which to be treated as x and which as y.
- Write the given information as two ordered pairs.
- Calculate the slope from the ordered pairs you constructed.
- Find the y-intercept using the linear model with the slope you found above entered and one of the data points.
- Put both the slope and y-intercept into the linear model.
Example 5
The number of words in a childs language is approximately linear when a child is between 12 months and 18 months (after about 18 months the growth in childrens vocabulary increases at a much faster rate). A particular childs vocabulary was tracked at age 12 months and age 18 months. At 12 months the vocabulary was 5 words and at age 18 months the vocabulary was 31 words. Create a linear model to show the relationship between number of words and age of a child.
Solution
To get started we should first identify the variables. We are given the age of a child and the number of words in the vocabulary. We will use the letter "a" to represent the age of the child and the letter "v" to represent the number of words in the vocabulary.
The age of a child best explains what we would see as a value for the number of words in the vocabulary. We will treat "a" as the x variable in the linear model and "v" as the y variable. This gives us our structure for the linear model as:
The information provide in the text gives us two data points in the form . At 12 months the child has a vocabulary of 5 words, so our first ordered pair is . The second ordered pair would be .
Calculate the slope:
Using the information about the slope and the ordered pair we can now find the y-intercept value:
The linear model that relates the age (in months), a, and the number of words in the vocabulary, v, is:
We found a linear model that related the childs age in months to the number of words in their vocabulary. What does the value of slope mean for that application? The slope in the above example was . When the slope was calculated the values used represented:
The slope was that change in number of words in the vocabulary over the change in age (in months). Putting this together with the value we have the interpretation of 13 words per 3 months. In other words we expect to see a gain of 13 words in the childs vocabulary for every three months of age.
What are the units of slope?
We have already looked at the formula for slope and saw it represents the when viewing it graphically or as the is looking at the formula of the form . In both situations what we have is the ratio of y values to x values. This means are units for slope will be the units of y over the units of x:
What about that y-intercept for our previous example? In this case the y-intercept would not meaning in the application as it would represent the number of words in the childs vocabulary at 0 months age. With what was told in the setup of the question a value of 0 for the age would not be okay to use for the linear model. In the next example we will examine the y-intercept in the application. In the work we will use to represent the time in years from 2019.
Example 6
In the Fall of 2019 the PCC tuition for an in-state residents was $84.50 for 1 credit. In Fall of 2020 the tuition for an in-state resident was $87.00 for 1 credit. Construct a linear model for the cost of in-state resident tuition of 1 credit hour assuming the trend is linear and continues in that pattern.
- What does the slope represent in our model?
- What is the predicted tuition in 2024?
- When would the tuition be over $150?
Solution
In this situation we have an independent variable of time (the years after 2019) and a dependent variable being the tuition cost as the year it is explains the cost in tuition we would see. Let t be the time (in years), but we will let represent the year 2019. The dependent variable being tuition cost can be represented by c.
The data that was given to use can be written as a coordinate pair . In 2019 the cost for tuition was $84.50, so the corresponding ordered pair is . In 2020 the tuition was $87.00 and gives the ordered pair as .
Find the slope based on the data provided:
In this instance we are given the y-intercept as we are using to represent the tuition in 2019 with a value of &84.50. This means the value of b, the y-intercept, is 84.50 for the model.
Putting the slope and intercept into the model we have:
- Our interpretation for the slope is that the tuition will increase by $2.50 per year after 2019.
The y-intercept value of 84.50 is the starting tuition amount in 2019. - To find the tuition in 2024 we will evaluate the model at as 2024 is fives years after 2019.
- To find when the tuition is greater than $150 we can set the tuition in the model to 150 and solve for t. The tuition reaches $150 a little after 26 years. If we use the model for and we see that it is over $150 when t=27 or the year 2046.
You may be asking why we wanted the start of the linear model in 2019 to be represented with . The reason for this is to help with reading the model after it is created and having a better understanding of where the tuition started from by looking at the y-intercept value. When the y-intercept is going to be $84.50; the starting tuition amount in 2019. If we had not picked that starting time as 2019 we would have ended up with a y-intercept that was negative and had not relation to the application for what happens when you enter 0 for the time (year).
Try it Now 4
The population of a small city was in 2012 was 32,500. By 2018 the city has grown to a population of 37,000. If we assume the population is growing with a linear trend, then find a linear model for the city's population and use the model to answer the following questions:
- What does the slope mean in the model?
- What would the predicted population be in 2021?
- In what year would the population reach 45,000?
Answer
Exercises
- Identify which are linear models. If it is a linear model state the slope and y-intercept.
Answer
- Find the equation of a line through the points
Answer
- Find the equation of a line through the points
Answer
- Find the equation of a line through
Answer
- Graph the linear model given by
Answer
- Graph the linear model given by
Answer
- Graph the linear model given by
Answer
- The population of beetles is assumed to be growing according to a linear model. The initial population was 30 beetles and after 8 weeks the population reached 62.
- What is a linear model to represent the amount of beetles after t weeks?
- How many beetles would we expect to find after 10 weeks?
- In what week would we expect to see the population reach 150?
Answer
- The population of a town was 120,000 in 2010. In 2020 the town population was 180,000.
- Write is a linear model to represent the population n years after 2010.
- What does the model predict the population to be in 2025?
- Based on the model when would the population reach 250,000?
Answer
- A student tutors to help pay for college expenses. If they charge a flat fee of $20 plus $15 an hour, then we can model the total cost of money charges by the following:
Describe what the slope and intercept represent in the model.
Answer
- A Manufacturer believes there is a linear relationship between the number of widgets, n, they can sell and the price, p, it can charge for the widget. They have historical data that shows they can sell 7,000 widgets at a price of $10 per widget and 1,000 widgets at a price of $25 per widget. Find a linear model that relates the price and number of widgets that can be sold.
Answer
- An internet provider charges for service based on the model:
, where n is the number of GB of data used and C is the monthly charge in dollars. Interpret both the slope and y-intercept for the model in terms of the application.
Answer