p-value for the unequal variance t-test is apparently wrong.

[Anonymous]

	TTEST_t	TTEST_df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	TTest_df(gender, Age, 0)	TTest_sig(gender, Age, 0)
Equal Variance not Assumed	-1.751	21.317094851827	0.172	-6.34127	3.62059	21.32	0.172
Equal Variance Assumed	-1.844	30	0.075	-6.34127	3.43822	21.32	0.172

p-value for the unequal variance t-test is apparently wrong. Comapre to Excel =TDIST(1.751,21.317095,2) or R:

>

t.test(Age~gender,data=age,var.eq=F,alternative ="two")

Welch Two Sample t-test

data: Age by gender

t = 1.7514, df = 21.317, p-value = 0.09425

Age	control	gender
159	0	F
155	1	F
151	1	M
147	1	F
146	1	F
144	0	F
142	0	F
142	0	M
141	1	M
139	1	F
138	1	M
137	0	F
136	1	F
135	0	F
135	1	M
132	0	M
131	0	M
130	1	M
129	0	M
129	1	F
129	1	M
123	0	M
123	1	M
122	1	M
121	1	F
116	0	F

[#2]

I stumbled across the same problem, appearently this has not been fixed