[EMAIL PROTECTED] (Katie) wrote in 
news:[EMAIL PROTECTED]:

> I have completed a survey (about tv) and I have some yes/no ranked
> data.  I also have groups, for example gender and age.
> 
> I'm completely unsure how to statistically analyse this.
> 
> From looking at my results, I can see in over 90% of males watched
> football, but only 65% of females did.
> 
> Is there a statistical test to compare these two groups and say they
> are statistically different, I have some other yes/no data that is
> closer together so a test would prove useful.

What are you asking?  Whether the proportion of football-watchers differs 
by gender?

> 
> I thought of a chi squared test (a 2x2 contigency one I think it was
> called) I tried it, but to be honest I was lost.  Couldnt find any
> useful examples on the internet and I think I will burn my stats book,
> becuase frankly it was useless at explaining chi squared tests.
> 
> Is the chi squared test the test to use to see if there is any
> difference between yes/no ranked data and 2 or more groups?  If so I
> would appreciate if someone could explain the jist or could point me
> towards a good example or something would be great.

If you're asking whether the proportion of football-watchers differs by 
gender, then a 2X2 chi-squared test is in fact the appropriate test.  A 2X2 
chi-squared test is pretty simple, and if your textbook can't adequately 
explain it, various sources on the Web, available through a Google search, 
will be able to do so.
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to