LU Interpreting Chi Square Invalid Hypothesis & P Values Response Discussion

You are replying to another students discussion Board post. The instructions for replying are as follows:

Review another students response to D5.1 and D5.2. Summarize their findings and indicate areas of agreement, disagreement and improvement. Support your view with citations and include a reference section.

Here is the students post you are replying to:

Discussion Forum 5

D5.1 Interpreting Chi-square

D5.1a

In output 8.1 (Pearson) chi-square is not statistically significant because p = 0.056. This p level is more than the preset alpha level probability of 0.05. According to Morgan et al (2013), if this p level was less than 0.05, it means the results are statistically significant and therefore the null would need to be rejected.

D5.1b

The expected values as depicted in output 8.1 are at least 80% and the cells are greater than 5 since the minimum is 14.05 as shown in the math grade* gender cross tabulation output and also stated in the footnote under the chi-square test output. It is important because it indicates that the data met the condition of at least 5 in each cell. Additionally, this satisfy the use of the two-sided chi-square test (Morgan et al., 2013).

D5.2 Measure Strength of the Relationship

D5.2a

Because “father’s education revised” and “mother’s education revised” are at least ordinal data, the most appropriate statistics to measure the strength of the relationship would be the Kendall’s tau-b because father’s education and mother’s education are ordered variables and ordinal data. According to Morgan et al (2013), if the Cramer’s V was used for these variables, it would have treated the cross tabulation as if they are nominal even if they are ordered, thereby resulting in an error in the statistics.

D5.2b

The two statistics are different because Kendall’s tau-b measures the strength of the association if both variables are ordinal while Cramer’s V measure the strength of the relationship of two nominal variables when one or more have three or more levels (Morgan et al, 2013).

References

Morgan, G., Leech, N., Gloeckner, G., Barrett, K. (2013). IBM SPSS for Introductory Statistics
(5th Ed.). New York, NY.