7 and 8, the first four estimated coefficients are subjected to a two-tailed test, whereas the last two coefficients are subjected to only a one-tailed test
And, all the independent variables will be tested with one-tailed tests
unless otherwise indicated.
Since this study uses a 5 percent p-value for a one-tailed test
, the rejection rate should be 5 percent.
However, if the alternate hypothesis was stated so that a one-tailed test
could be used (that is, male lawyers make more money than female lawyers), it is entirely conceivable that the one-tailed test
could find that a $2,000 difference is statistically significant at a P value less than .
Indicates statistical significance at the 1% level, one-tailed test
where z is a factor for a one-tailed test
of significance (usually set at 1.
For the variables that aggregate positions of villagers and township leaders, I use a one-tailed test
01 level for a one-tailed test
and a two-tailed test.
In fact, if one were to consider a one-tailed test
instead of the two-tailed test reported, a Z of -1.
Numbers in parentheses are the critical t-values at a 5% significance level for a one-tailed test
In the present study the findings occurred with use of a one-tailed test
and at a borderline level for females.
The use of a one-tailed test
, of course, is invalid when the hypothesis tested is itself derived from the data set that is being tested.