The name is absent

Table 1

Effect of Observation on Induction of Placcntophagia and on Maternal Sensitization Latency
in Nonplacentophagic, Nonmaternal Virgin Rats

*Access groups significantly greater than по-access groups (p < . 001).

↑Signiftcantly greater than value for chow∕access group (p < . 05).

P = .62, Fisher exact probability test). Therefore, ob-
serving a rat eating chow was as ineffective in altering
the incidence of placentophagia as was observing a
rat that was not eating.

A chi-square analysis was applied to the differences
among the three observation conditions for the ac-
cess group only (part∕access, plac∕access, and chow/
access). Chi-square rather than ANOVA was used
because in cases involving small groups and dichot-
omous data, the chi-square test is more powerful
(Narula & Levy, 1977). The groups in the no-access
condition were analyzed separately because the ex-
pected frequencies were too low for analysis with chi-
square.

The analysis, using an expected frequency of 5
computed from the proportion of pɪaeentophages
produced in the baseline condition, indicated that
there were significant differences among the obser-
vation conditions [χ²(2) = 6.8, p < .05]. Three subse-
quent two-way chi-square tests, which were adjusted
for a joint alpha level with Bonferroni’s procedure
(Myers, 1972), indicated that the proportions of pla-
Centophages in the part∕access and plac∕access groups
were significantly greater than that in the chow∕access
group, but were not significantly different from each
other (joint a = .05).

The observation conditions on the no-access group
were not significantly different from each other (3/15
vs. 1/15; p= .30, Fisher exact probability test).

A 3 × 2 ANOVA, comparing three levels of ob-
servation and two levels of access to placenta, was
used solely to determine whether there was any effect
of the interaction between the two variables. The re-
sults indicated that there was no significant effect of
this interaction on placentophagia in Test 1 [F(2,84)
< 1.0].

The difference between the access and no-access
conditions was analyzed by pooling the data across
observation groups and computing the standard
error of the difference between the two resulting pro-
portions, 23/45 and 5/45 (Dixon & Massey, 1969).
The proportion of placentophages in the access group
was significantly greater than that in the no-access
group (dp,-p_j = .089, z = 4.09, p < .001).

Maternal Sensitization Latency

The mean latencies (± SEM), in days, for the onset
of maternal behavior during the concaveation proce-
dure (maternal sensitization latencies) for the groups,
are presented in Table 1.

An ANOVA was used to determine whether there
was any effect of the three types of observation or of
the two access conditions on the rate of onset of pup-
induced maternal behavior. The results of the analy-
sis indicated that there was no significant effect of
observation condition [F(2,84) = 1.80, p > .05], of
access condition [F(l,84) < 1.0], or of the interaction
between those two main effects [F(2,84) = 1.40,
p > .05].

Test 2

The Test 2 data consisted of the results of a placen-
tophagia test conducted after the Concaveation pro-
cedure.

An ANOVA, which compared two tests (repeated
measures), three observation conditions, and two ac-
cess conditions, indicated that there was no signifi-
cant difference between Test 1 and Test 2 [F(l,84)
= 1.04, p > .05]. The interaction between the access
variable and the test variable was significant [F(l,84)
= 4.15, p= .04]. This was because one pɪaeentophage
in the part∕access group and two in the plac∕access
group became placentophagic after the maternal sen-
sitization phase of the experiment (see Table 1).
Also, one rat in the part∕no-access group that had
eaten in Test 1 did not eat in Test 2. It is impossible

<<   <   1   2   3   4   5   6   >   >>