4 Concluding remarks

Table 8 summarizes the most salient results. The mean circumference resembles very closely the mean circumferences of 120 mm reported by Waldeyer in 1889, by Delbet in 1901, and by Testus in 1931 (Dickinson, 1933/1971; see Table 1). Moreover, this value equals the median circumference of 121 mm derived from the data of Kinsey and colleagues (Kinsey et al., 1948; see Introduction). The P25, P75, and P90 approximate the P-values made up from the frequency tables of Kinsey and collegues (Gebhard & Johnson, 1979). The P25 derived from their data (N = 2505, maximum = 171 mm) lies between 108-114 mm, the P75 between 127-133 mm, and the P90 between 140-146 mm.

Table 8. Summary for subjects with 'full erection' and responses/NPT of more than 10 mm

Criterion	N	mean	median	sd	range	P25*	P75	P90
Full erection	45	122	118	18	90-160	109	132	154
Response > 10 mm	74	122	121	16	90-161	110	131	142
NPT > 10 mm Total	25 144	116 121	116 119	10 16	90-135 90-161	109 110	125 130	128 141

*P25, P75, and P90 describe the circumference value (mm) that encompasses 25%, 75% and 90% of the subjects, respectively. (The median describes the circumference value that encompasses 50% of the subjects).

Some final remarks concerning the validity and generalizibility of the data are in place. First of all, it should be noted that the circumference values are based on self-reports. Subjects were asked to obtain their own readings of the Erectiometer. This procedure may increase the probability of inaccuracies. As can be seen in Table 3, circumference values obtained in study 4 were substantially higher than in the other three studies. It is uncertain to what extend this was effected by measurement inaccuracies or by other factors such as differences in subject groups or stimuli. Secondly, the Erectiometer was positioned approximately half-way down the penis. This suggests, together with the fact that the Erectiometer requires force to expand, that circumference of the base of the penis may have been higher, at least to some degree, in a subgroup of subjects. Thirdly, the criterion for 'full erection' was not determined independently, but was, instead, defined by the subjects themselves. The second index used for 'degree of erection' was circumference change. While circumference change may be a reasonably accurate predictor of rigidity (degree of erection), individual differences in the amount of change constituting full erection are most likely large.

In sum, the presented data should be interpreted with some caution. Certainly, some of the methodological pitfalls or problems mentioned are general to this kind of research, and it may also be clear that any solution or approach has its pros and cons. Still, while this study has its own drawbacks, we believe that the presented data delineate a realistic range of penile circumferences, with values closely resembling values previously reported in other countries.