In this study, they speculate that perhaps intercourse timing may seem to be correlated to gender ratio because of an actual relationship of gender ratio to length of the follicular phase.
In 1979, a scientist named Harlan studied a group of married Orthodox Jews that achieved conception. She studied married Jewish women in Israel who
practised abstention from sexual intercourse for 7 days after the cessation of menstrual bleeding, as required by orthodox ritual. Date of first intercourse was inferred from the date of the wife's visit to the ritual bath (mikve), presumed to mark the resumption of marital relations. She found a statistically significant relationship between the timing of the mikve visit and the proportion of boys born, with an excess of boys among cycles where the ritual bath was 2 days after presumed ovulation.
While Harlap (1979) interpreted this as suggesting an effect of timing of intercourse vis-a-vis ovulation, her data are open to a different interpretation. Lacking a direct marker, she inferred the date of ovulation in conception cycles by subtracting 14 days from the woman's usual cycle length. This approach relies on the fact that the luteal phase of the cycle is relatively fixed, lasting -13-14 days. However, the group characterized as having a late visit to the mikve must have predominantly been those whose usual menstrual cycle was short, i.e. those who had a short follicular phase. For women with a typical menses length of 5 days, intercourse could resume on cycle day 12, according to ritual. A woman whose usual cycle length is 24 days would usually ovulate around cycle day 10, 2 days before the ritual bath. In this way, women judged to have visited the mikve 2 days after ovulation must have predominantly been those with short cycles and short follicular phases. It was only in this particular small group that Harlap (1979) found an unusual sex distribution. The high male to female ratio seen in these women may have been due to the preponderance of short follicular phases in this group,
and only incidentally related to the timing of intercourse relative to their estimated day of ovulation. If there is a relationship between the timing of intercourse in relation to ovulation and the sex of the baby, then our finding could porentially be explained in part by a relationship between the patterns of intercourse and the length of the follicular phase. We observed no such relationship. All couples in our study were trying to conceive, and in most cycles intercourse had been reported close to the time of ovulation and earlier as well. Pregnancy could rarely be attributed to a single event. During the fertile interval consisting of the 6 days up to and including the day of ovulation (Weinberg et ai, 1994), intercourse was reported on an average of 2.5 of the days, with no difference in this total for cycles in which boys versus girls were conceived. Moreover in our data, closeness of intercourse to day of ovulation was not related to the sex of the baby.