- The curve is nonsymmetrical and skewed to the right.
- There is a different chi-square curve for each dfdf.
- The test statistic for any test is always greater than or equal to zero.
- When
df
>
90
df>90, the chi-square curve approximates the normal. For
XX ~
χ
1000
2
χ
1000
2
the mean,
μ
=
df
=
1000
μ=df=1000
and the standard deviation,
σ
=
2
⋅
1000
=
44.7
σ=
2
⋅
1000
=44.7.
Therefore,
XX ~
N
(
1000
,
44.7
)
N(1000,44.7), approximately.
- The mean,
μ
μ, is located just to the right of the peak.
In the next sections, you will learn about four different
applications of the Chi-Square Distribution. These hypothesis tests are
almost always right-tailed tests. In order to understand why the tests are
mostly right-tailed, you will need to look carefully at the actual
definition of the test statistic. Think about the following while you
study the next four sections. If the expected and observed values are
"far" apart, then the test statistic will be "large" and we will reject in
the right tail. The only way to obtain a test statistic very close to
zero, would be if the observed and expected values are very, very close to
each other. A left-tailed test could be used to determine if the fit were
"too good." A "too good" fit might occur if data had been manipulated or
invented. Think about the implications of right-tailed versus left-tailed
hypothesis tests as you learn the applications of the Chi-Square
Distribution.
"Reviewer's Comments: 'I recommend this book. Overall, the chapters are very readable and the material presented is consistent and appropriate for the course. A wide range of exercises introduces […]"