Testing group differences in sequential designs is based on critical values of the standard normal distribution. This approximation is not called into question in various variants if sequential tests. It is the basis not only for assessment of the null hypothesis but also for calculating sample size and power. Additional assumption concerning this kind of tests are homogeneity of variances and normal distributed random variables. In a simulation study effects of using the α-percentile of the standard normal distribution and of violating various prerequisites on type I error rate and power are evaluated. Although the focus lies on triangular tests based on Whiteheads method [1] for continuous monitoring other sequential designs like the group sequential test of OBrien-Fleming [2] and Pocock [3] were also taken into consideration.

1.
J.
Whitehead
and
I.
Stratton
,
Biometrics
39
, p.
227
236
(
1983
).
2.
P. C.
OBrien
and
T. R.
Fleming
,
Biometrics
35
, p.
549
556
(
1979
).
3.
S. J.
Pocock
,
Biometrika
64
,
191
199
(
1977
).
4.
A.
Wald
,
The Annals of Mathematical Statistics
16
,
117
186
(
1945
).
5.
P.
Armitage
,
C. K.
McPherson
, and
B. C.
Rowe
,
Journal of the Royal Statistical Society, Series A
132
,
235
244
(
1969
).
6.
A. G. R. L. D.
Rasch
,
J.
Pilz
and
Verdooren
, “Optimal experimental design with r,” in
Sequential Designs
, edited by
B.
Raton
(
CRC Press, Taylor & Francis group
,
2011
), pp.
139
169
.
7.
A. I.
Fleishman
,
Psychometrika
43
(
4
),
521
532
(
1978
).
8.
S. I.
Inc
.,
SAS/STAT 9.2 Users Guide
(
SAS Institute Inc., Cary, NC: SAS Institute
,
2008
), pp.
25
30
.
9.
J. M.
Kittelson
and
S. S. S. S.
Emerson
,
Biometrics
55
, p.
874882
(
1999
).
This content is only available via PDF.
You do not currently have access to this content.