
1. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 1, 2013 6:59 AM (in response to canny yue)HI Cunny,
I suggest you to read this thread: Using table calculation to calculate chisquared where Scott Tennican provides a solution and a sample workbook with implementation of chisquared in contingency table.
Best
Ramon

2. Re: is possible base on my data and get chisquare result？
canny yue Nov 3, 2013 5:05 PM (in response to Ramon Martinez)Hi Ramon,
i know that post, but i don't understand their working, and i follow their formula what does the output for?
thanks
canny

3. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 3, 2013 7:47 PM (in response to canny yue)1 of 1 people found this helpfulHi Canny,
The sheet TwoCategoricals in the workbook chiSquaredMultiPopTest.twbx created by Scott Tennican shows a TwoWay Table (or cross tab) that counts the number of subjects (in this specific case is the number of sales) by Product Category and Region, including marginal counts that you can see in column and rows grand totals.
When you hover the mouse over every cell of the table, the tooltip shows the Counts of Sales, Count of sales Expected and the value of ChiSquared that is 12.25
Now we need to determine the degree of freedom for the TwoWay Table using the formula
df=(number of rows  1)*(number of column  1)
In this specific table df=6
Having ChiSquared=12.25 and df=6 we need to go to the ChiSquared distribution table to determine the probability or Pvalue P(ChiSquared >= 12.25) for chisquared distribution with 6 degree of freedom
Pvalue is 0.056
As pvalues is higher that 0.01, we can concludes that there is no evidence of association among Sales by Region and the factor Categories of Products.
There is a drawback in the solution: the degree of freedom and the pvalue is not calculated, so you need to calculate df by yourself and find the pvalue in a ChiSquared distribution table.
The web page "Calculation for the ChiSquared Test" provides an online tool to calculate chisquared, df and pvalue
References
 TwoWay Tables and the ChiSquared Test http://www.stat.yale.edu/Courses/199798/101/chisq.htm
 ChiSquared Distribution Table http://sites.stat.psu.edu/~mga/401/tables/Chisquaretable.pdf
 Calculation for the ChiSquared Test http://www.quantpsy.org/chisq/chisq.htm
I hope this helps you to understand the output of the twoway table and chisquared test.
Best,
Ramon

4. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 3, 2013 8:14 PM (in response to canny yue)This video can help to understand ChiSquared Test
Best,
Ramon

5. Re: is possible base on my data and get chisquare result？
canny yue Nov 3, 2013 9:03 PM (in response to Ramon Martinez)Hi Ramon,
how you know P value is 0.056 and df=6?
thanks
canny

6. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 3, 2013 9:32 PM (in response to canny yue)Hi Canny
To calculate df count the number of rows and number of columns in the twoway table and use formula bellow
Df=(rows1)*(columns1)
Go to chisquare distribution table and find the critical value of chisquared based on df and 95% confidence.
If your calculated chisquared is lower than the critical value, reject the null hypothesis.
I fond the pvalue in the online tool I've mentioned in my previous comment.
Ramon

7. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 5, 2013 12:10 AM (in response to canny yue)Hi Canny,
Have you done any progress in understanding the Towway Table and Chisquared test?
Let me know if you require additional support.
Regards,
Ramon

8. Re: is possible base on my data and get chisquare result？
canny yue Nov 5, 2013 12:31 AM (in response to Ramon Martinez)Hi Ramon，
i understand the Chisquared test. but i still cannot get the similar result like beginning i post that picture.
Thanks
canny

9. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 5, 2013 6:29 AM (in response to canny yue)Hi Canny,
Could your explain better what you need to accomplish? It is also appreciated if you explain the content of the screenshot you you included in your first post.
It seems it represents a kind of Twoway table but I don't see any result on it.
Best,
Ramon

10. Re: is possible base on my data and get chisquare result？
canny yue Nov 5, 2013 4:58 PM (in response to Ramon Martinez) 
11. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 5, 2013 5:03 PM (in response to canny yue)Hi Cunny,
What is the software that produce that results. To tell you the truth, I don't see any ChiSquare result in the screenshot you have provided.
Ramon

12. Re: is possible base on my data and get chisquare result？
canny yue Nov 5, 2013 5:07 PM (in response to Ramon Martinez)Hi Ramon,
i use SPSS modeler ,there also did not show me the working(but i know that is using chisquare), is a device that can automatic calculate the result.
thanks
canny

13. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 5, 2013 6:05 PM (in response to canny yue)HI Canny,
I don't have access to SPSS as I'm not in my office right now. I can check tomorrow in the morning.
But to tell you the truth any ChiSquared method has to provide the following elements:
1. Twoway table with counting by cells and grand totals by columns and rows
2. ChiSquare value
3. Degree of freedom
4. Pvalue
I don't see those elements in the results (screenshot) you got in SPSS
Best,
Ramon

14. Re: is possible base on my data and get chisquare result？
Ramon Martinez Nov 6, 2013 12:46 PM (in response to canny yue)Hi Canny,
I applied Chisquared to a data set using SPSS and this is the output. I used SPSS ver 17.
First you are able to see the Twoway table crossing to dimensions: Know to read and write vs. ***. Bellow the twoway table are the summary statistics: Pearson ChiSquate, df and Pvalue
I hope this help.
Regards,
Ramon