01.03.2013 Views

Applied Statistics Using SPSS, STATISTICA, MATLAB and R

Applied Statistics Using SPSS, STATISTICA, MATLAB and R

Applied Statistics Using SPSS, STATISTICA, MATLAB and R

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

182 5 Non-Parametric Tests of Hypotheses<br />

these intervals, under the “z-Interval” heading, which can be obtained from the<br />

tables of the st<strong>and</strong>ard normal distribution or using software functions, such as the<br />

ones already described for <strong>SPSS</strong>, <strong>STATISTICA</strong>, <strong>MATLAB</strong> <strong>and</strong> R.<br />

The corresponding interval cutpoints, xcut, for the r<strong>and</strong>om variable under<br />

analysis, X, can now be easily determined, using:<br />

xcut cut<br />

= x + z s , 5.9<br />

X<br />

where we use the sample mean <strong>and</strong> st<strong>and</strong>ard deviation as well as the cutpoints<br />

determined for the normal distribution, zcut. In the present case, the mean <strong>and</strong><br />

st<strong>and</strong>ard deviation are 137 <strong>and</strong> 43, respectively, which leads to the intervals under<br />

the “ART-Interval” heading.<br />

The absolute frequency columns are now easily computed. With <strong>SPSS</strong>,<br />

<strong>STATISTICA</strong> <strong>and</strong> R we now obtain the value of χ *2 = 2.2. We must be careful,<br />

however, when obtaining the corresponding significance in this application of the<br />

chi-square test. The problem is that now we do not have df = k – 1 degrees of<br />

freedom, but df = k – 1 – np, where np is the number of parameters computed from<br />

the sample. In our case, we derived the interval boundaries using the sample mean<br />

<strong>and</strong> sample st<strong>and</strong>ard deviation, i.e., we lost two degrees of freedom. Therefore, we<br />

have to compute the probability using df = 5 – 1 – 2 = 2 degrees of freedom, or<br />

equivalently, compute the critical region boundary as:<br />

2<br />

2 , 0.<br />

95<br />

χ = 5.<br />

99 .<br />

Since the computed value of the χ *2 is smaller than this critical region boundary,<br />

we do not reject at 5% significance level the null hypothesis of variable ART being<br />

normally distributed.<br />

Table 5.7. Observed <strong>and</strong> expected (under the normality assumption) absolute<br />

frequencies, for variable ART of the cork-stopper dataset.<br />

Cat. z-Interval Cumulative p ART-Interval<br />

Expected<br />

Frequencies<br />

Observed<br />

Frequencies<br />

1 ]− ∞, −0.8416] 0.20 [0, 101] 10 10<br />

2 ]−0.8416,<br />

−0.2533] 0.40 ]101, 126] 10 8<br />

3 ]−0.2533,<br />

0.2533] 0.60 ]126, 148] 10 14<br />

4 ] 0.2533, 0.8416] 0.80 ]148, 173] 10 9<br />

5 ] 0.8416, +<br />

∞ [ 1.00 > 173 10 9

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!