turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- How to cross tabulate statistical significance usi...

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 05:46 AM

Hi,

I have a situation where i have values for products in 1,0.

I want to compare all vs all products and need chisq statistics for them.

like for A vs B, B vs C. and statistical significance for all of those values?

Thanks

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 09:43 AM

Check out the documentation for PROC FREQ and its TABLES statement.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 10:08 AM

The freq and tables statement though computes the frequencies of 1,0 for each product. I want the CHISQ statistic for all product vs product comparison. If I run at one time, it give only overall CHISQ

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 10:50 AM

Post a sample of your data. Any reason you can't do comparisons selectively or are you trying to adjust your p-values?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 10:58 AM

the data would be like:

PRODUCT | VALUE | |

A | 1 | |

A | 0 | |

A | 0 | |

A | 1 | |

A | 1 | |

B | 0 | |

B | 0 | |

B | 1 | |

C | 0 | |

C | 1 | |

C | 0 | |

C | 1 |

I want comparison of A vs B, A vs C and C vs A. One method is subsetting data for each one. But I want to do it without subsetting similar to ANOVA.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 12:15 PM

I think Proc CatMod is what you're looking for, but not sure it does chisquare analysis.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

04-28-2014 12:24 PM

well, i had tried logistic regression. But the p values, are actually the probability of intercept being 0/1 for that particular variable, and not for statistical significance of CHISQ