Chi-square test - I do not understand the results :-&

[markecb]

In one case I get the results as a significant difference in the other with the same dana, the difference is not significant.


Data sample
A B
I 5 8
II 4 28

http://imageshack.com/a/img542/467/o61k.jpg

Same data and got significant difference

http://imageshack.com/a/img200/7366/y1le.jpg

Where I'm wrong?

 

[Stephen Tashi]

The images don't say what kind of chi-square test is being done.

Try using various options for chi-square tests on the page http://graphpad.com/quickcalcs/contingency1/. Different statistical tests on the same data may give different conclusions.

 

[FactChecker]

The second one says it does not apply a Yates correction. It's results are just barely significant at the 95% level. The first test may apply the correction and the result might become insignificant or it may use a different level than 95%.

 

[markecb]

Thank you, here is little more data; I used chi-square (χ2) test to investigate the statistical significance of differences of opinion about the students offered the choice of extra-curricular activities with regard to gender and school achievement, ie the most recently completed assessment of students in the English language.

Can I (completed by SPSS) results interpreted as: no statistically significant differences based on gender.?

http://imageshack.com/a/img837/3372/0wv6.jpg

 

[FactChecker]

The test tried to prove that there was a difference with 95% confidence. It barely succeeded by Pearson chi-squared and barely failed by the other statistics tests. You should NOT conclude that it proved there was no difference. It was not trying to prove that.