Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
41 views

Linear Modeling

The document discusses a statistics class activity on linear regression. Students observed a dataset of commute times and well-being scores from a survey of American adults. They found a strong negative correlation between commute time and well-being, with longer commutes associated with lower well-being scores. The regression line equation allowed them to predict well-being for different commute times, such as a score with a 120 minute commute. Overall, the activity demonstrated how to use linear regression to model relationships in data and make predictions.

Uploaded by

api-242750497
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
41 views

Linear Modeling

The document discusses a statistics class activity on linear regression. Students observed a dataset of commute times and well-being scores from a survey of American adults. They found a strong negative correlation between commute time and well-being, with longer commutes associated with lower well-being scores. The regression line equation allowed them to predict well-being for different commute times, such as a score with a 120 minute commute. Overall, the activity demonstrated how to use linear regression to model relationships in data and make predictions.

Uploaded by

api-242750497
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Andrew Belov

Professor Julie Humphreys

Math 1040 Statistics

5/3/2017

Linear Modeling

In our first activity, we learned about linear regression. The linear

regression model is a method in which we may observe and conclude if a

specific set of data presents a negative or positive linear relationship. The

data we observe comes from a scatter plot or a graph. The variables in the

data set are referred to as explanatory (X), and response (Y). We use the

formula for the linear correlation coefficient (r), and compare the

absolute value of (r), to a critical value, which is based on a sample size. If

the r value is larger than the critical value, we can then determine if a

correlation exists. The closer the (r) value is to 1 or -1, the correlation is said

to be strong positive, or a strong negative. From here we calculate the

coefficient of determination (r2). This value is expressed as a percentage,

and explains the variance in X, can be explained by Y. We went on to find a

least squares regression line, dictated by y hat. (y^=ax+b) This regression

line gives us the equation of our data. We use this Y^ formula to make

predictions regarding our variables.


In our activity, we looked at a survey of adult Americans regarding

their commute times to work as well as their well-being score (given by a

survey). Our Explanatory variable was commute time (in 5,10, 15... mins

etc.) Our Response variable was the corresponding well-being score. We

found that with a sample size (n)=10, the corresponding critical value was

0.632 We compared the (r) value, and found that a strong negative

correlation existed. Our r2 value came out to be .8116, meaning, 81.16% of

the variation in y (well-being score) can be explained by x (commute time).

The longer the commute time, the worse the score becomes. We then used

the y hat equation and made a prediction for a well-being score with a

commute time of 120 minutes. The well-being score ended up with a low

value.

This method taught us how to make predictions about data. As long as

the data has a linear relationship, and is within your data set, you can make

predictions on what to expect on average, for just about anything.

You might also like