Discussion
Topics
Last Day
Last Week
Tree View

Documentation
Getting Started
FAQ
Formatting
Troubleshooting
Program Credits

Utilities
New Messages
Keyword Search
Contact
User Profile
Administration

 
Guessing Correlations

intro.psych (Psyco 105) Discussion: Group 3 Discussion Group: Methods: Sites to Visit: Guessing Correlations
By
Connie Varnhagen (Admin) on Friday, August 28, 1998 - 04:08 pm:

What makes a correlation difficult or easy to guess? Discuss properties of correlations. What does a correlation between two variables mean? What types of research questions are correlations most appropriate for answering?
Go There


By Admin on Wednesday, September 9, 1998 - 07:05 pm:

Hey guys!

Not all the correlations were correct (well, maybe I'm being presumptuous) when I did this site lately.

Connie Varnhagen


By Montresor on Thursday, September 10, 1998 - 03:30 pm:

I tried it!!! My second time I got 6th place (Score of 80 in a row)

Graphically, correlations are skinnier the higher the correlation. The sign depends on the direction of the slope.
I realize now how correlations can be easy to guess. But sometimes you get the real hard ones, the ones that vary in only a percent or so and you have to almost guess on those ones.


By Jene on Sunday, September 13, 1998 - 11:11 am:

It is easy to guess a correlation if the data on the plot is clustered closely together and slanting upwards (positive correlation) or downwards (negative correlation). It is more difficult to guess a correlation if the data is spread out or scattered.

Properties of correlations: A positive correlation is one in which an increase in one variable leads to an increase in a second variable. A negative correlation is one in which an increase in one variable leads to a decrease in a second variable. A correlation coefficient measures the strength of the correlation (from +1.00 to -1.00). Did I answer that question right?

A correlation between two variables measures whether or not one variable can be used to predict another variable.

Correlations are appropriate for research questions that are measured numerically.

P.S. Who is writing the summary? Is this the topic we are choosing as a group?


By Admin on Sunday, September 13, 1998 - 12:35 pm:

Jene and Motresor:

Your responses are both excellent.

Although we often just talk about statistics, it is essential to look at the scatterplot to understand you data first before calculating the r. As you saw in the module, there are lot of different ways to get a particular correlation.

Connie V


By Mollyc on Monday, September 14, 1998 - 02:02 pm:

I cannot get any graphics, or to the correlation study there is a problem with the hook up. I tried earlier today and I will try again. And then send another message.


By Admin on Monday, September 14, 1998 - 02:24 pm:

Mollyc,
did you click on get new graphs? I had to do that. What version of Netscape or Internet Explorer were you using? Both must be 2.0 or higher, I think.
Connie V


Add a Message


This is a private posting area. A valid username and password combination is required to post messages to this discussion.
Username:  
Password:
Post as "Anonymous"