Chi-Square measurements can tell us whether or not a relationship is statistical

Question

Chi-Square measurements can tell us whether or not a relationship is statistically significant but not wether or not it is substantively related. So, if we find a Chi-square value of 57.33, what can we say?
a) the difference between the variables is primarily due to sampling error
b) the relationship between the variables exists and isn't due to sampling error but could be too small to be relevant
c) no relationship exists between the two variables but there may be a control variable that influences both of them
d) we must have calculated wrong because Chi-Square cannot be higher than 1

Explanation / Answer

A chi^2 value of 57.33 is large, and usually has a small P value (unless you have so many degrees of freedom, which is unlikely).

However, even if the P value is significant, there may be very small practical effects, and chi^2 cannot measure this.

So,

OPTION B: b) the relationship between the variables exists and isn't due to sampling error but could be too small to be relevant [ANSWER, B]

Related Questions

Navigate

• Browse (All)
• Browse C
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.