Questions on Finger-length Ratio and Personality

Several articles such as this one are floating about just now, but all of them leave open a couple of obvious follow-up questions.

To summarize: researchers have found that the ratio of the lengths of one’s index finger to one’s ring finger is affected by the amount of testosterone one was exposed to while still in the womb.

For women, “normal” is for the two fingers to be about the same length - for men it’s for the ring finger to be slightly longer. But for those who were exposed to higher testosterone levels before birth, the ring finger will be somewhat longer than “normal” for their sex.

Furthermore, they’re also finding that a number of less-trivial aspects of a person at adulthood are also related to this. The current articles discuss the finding that there are higher aggression levels among people exposed to higher testosterone in the womb, but previous studies also point to links with fertility, attractiveness to the opposite sex, sexual preference, tendencies toward sex-linked diseases, etc.

The upshot is that this finger-length ratio can be a simple physical marker that relates statistically to important aspects of ourselves.

Anyway, as you can tell I’ve been through the articles in some depth, but still haven’t seen the answers to two important questions:

  • What are the important ratio values? How is the measurement done, and what constitutes a “significant” variation from “normal” for your sex?

  • What factors determine testosterone levels in the womb? Would, for instance, emotional stress on the part of the mother during pregnancy affect hormone levels?

Does anyone know any more details on this?

The best information is probably available from the source (PDF). In summary: Finger lengths are measured from the basal crease (which I assume is the crease at the base of the finger on the palm side) to the fingertip. From Table 1, the mean(std.dev.) ratios were 0.947(0.029) for men and 0.965(0.026) for women. Figure 1 is a rather-unconvincing plot showing the correlation between the ratio and physical-aggression scores. The multiple-regression data (Table 3) may be more convincing (it would be hard for it to be less convincing), but I haven’t looked at it.

The sample size is pretty small. 149 males and 149 females, of only university students is a skinny sample from which to make such sweeping generalizations. Has anyone been able to duplicate the results?