Chem. Senses 27: 681-689,
2002
© Oxford University Press 2002
Gender Differences in the Retention of Swahili Names for Unfamiliar Odors
Department of Psychology, Macquarie University, Australia
Correspondence to be sent to: R.J. Stevenson, Department of Psychology, Macquarie University, NSW 2109, Australia. e-mail: rstevens{at}psy.mq.edu.au
 |
Abstract |
|---|
 Top Abstract IntroductionExperiment 1Experiment 2General discussionReferences |
|---|
Several studies, using different techniques, have established that women typically outperform men in naming odors. The mechanism for this effect was explored here in two experiments. In experiment 1, men and women learned randomly assigned Swahili names for a set of seven unfamiliar odors. Following multiple acquisition trials, participants were retested 1 week later. Although learning rates were identical during acquisition, after the 1 week interval, females were able to name more of the odors than men. Experiment 2 used a similar design but also included a retroactive interference task following the 1 week retention interval test. Although the week-long interval had the same effect as in experiment 1, interference had no effect on male or female performance. These results suggest that under conditions where experience is equated, female naming advantage may result from better consolidation of the learned material.
|
Introduction |
|---|
|
TopAbstractIntroductionExperiment 1Experiment 2General discussionReferences |
|---|
Correctly naming an odor involves at least two processes, recognizing the odorant and retrieving its label. Naming ability has been shown to systematically differ between males and females, with females consistently better at naming odors under a variety of different conditions (Doty, 2001
). Explanatory
accounts of this gender difference can be broadly divided into those based
upon experiential differences in olfaction and biological accounts, based upon
innate differences in olfactory perception and/or verbal ability. These
possibilities are explored here in two experiments, both of which use a paired
associate learning procedure to determine whether male and female participants
differ in acquisition, retention and proneness to interference of unfamiliar
odor—name pairs.
Gender differences in naming odors have been observed using a variety of
different procedures. (i) Several studies have employed the University of
Pennsylvania Smell Identification Test (UPSIT), in which participants are
asked to scratch and sniff an odor and then pick its name from a list of four
alternatives. Using this test, a female naming advantage has been observed at
all ages from 5 to 80 (Doty et
al., 1984; Gilbert and
Wysocki, 1987; Ship et
al., 1996), in individuals who have suffered damage to their
sense of smell (Deems et al.,
1991), between male and female twins (Segal et al., 1993)
and between males and females from a variety of different cultures
(Doty et al., 1985).
(ii) Differences in naming ability have also been observed under conditions
where odors are presented with no list of names
(Cain, 1982;
Engen, 1987). Effects in such
studies are of a similar magnitude to those observed using the UPSIT. (iii)
The National Geographic smell survey, completed by 1.5 million respondents
(Gilbert and Wysocki, 1987),
involved scratching and sniffing six odors and selecting the most appropriate
label from a list of alternatives. For each of the six odors, women
outperformed men on naming. (iv) Women have also been found to be better at
identifying the source (a process akin to naming) of biologically relevant
odors such as sweat (Wallace,
1977; Schleidt,
1980; Schleidt et
al., 1981). Thus, overall, evidence from a range of
procedures suggests a female superiority in odor naming.
The origin of this difference in naming ability between men and women can
be fundamentally attributed to two major causes, experiential and biological.
From an experiential perspective, differences between genders may emerge due
to societal influences on odor exposure
(Cain, 1980). For example,
women, who may be more likely to prepare food and use scented products, could
have a greater interest and motivation to learn the names of such odors. This
might ultimately translate into a heightened interest in all olfactory
stimuli, a view consistent with the findings from two recent surveys. First,
women's dreams appear to contain significantly more reference to odors than
men's (Zadra et al.,
1998). Second, women claim that smell plays a significantly
greater role in their choice of sexual partner than men
(Herz and Cahill, 1997). If
women do pay more attention to everyday odors than men, this would result in
an inevitable confound in most naming experiments. This is because naming
studies, by necessity, must use familiar everyday test odors (how could you
know the name of an unfamiliar odor?). It is, therefore, not currently
possible to exclude experiential factors as an explanation of gender
differences in naming.
Biological accounts have received far greater attention and can be divided
into three classes of explanation. In the first, it has been suggested that
the superior verbal ability of women may somehow account for better odor
naming (Engen, 1987). This
explanation is difficult to sustain, as recent meta-analytical data suggests
that male—female differences in verbal ability are very small, with an
average difference in ability of 0.1 SD
(Hyde and Linn, 1988;
Hyde and McKinley, 1997). In
the light of this and other recent findings
(Feingold, 1992;
Hedges and Nowell, 1995), it
would appear that general differences in verbal ability probably have
little part to play in explaining female superiority in odor naming. However,
it is plausible that there are specific differences in verbally
related abilities. For example, women may be better at forming associative
links between configural stimuli such as odors and faces and their verbal
label. Indeed, women are better able to recall the name for a particular face
than men (Witryol and Kaess,
1957; Thakur et al.,
1981).
A second type of biological explanation is based upon the much better
supported notion that women are usually better at all olfactory tasks than men
(Engen, 1987). One possible
source for such effects is the menstrual cycle, as it is well established that
women's odor sensitivity fluctuates over its course (Synder and Wolf, 1955;
Doty et al., 1981).
However, it is unlikely that variations in sensitivity resulting from the
menstrual cycle play any role in the enhanced ability of women to name
suprathreshold odors. First, in all the naming studies we are aware of,
menstrual cycle was a random variable, yet female naming advantage was
consistently observed. Second, female advantage in naming is evident prior to
puberty (i.e. before menstruation starts) and is present after
menopause, without any obvious reduction in effect size
(Doty et al., 1984;
Gilbert and Wysocki, 1987).
This is not to suggest that the menstrual cycle has no effect on naming, but
merely that female superiority in naming, as currently observed, is unlikely
to result from this cause.
A third biological explanation for gender differences in naming could arise
from some difference in brain structure or function
(Nopoulos and Andreasen, 1999;
Swaab et al., 2001).
At its simplest, this might mean enhanced olfactory receptor mechanisms in
females. This would presumably suggest generalized benefits in olfactory
sensitivity, as have been observed for many odors where menstruation is a
random variable (Koelega and Koster,
1974). Note, however, such effects have not manifested in all such
studies (Venstrom and Amoore,
1968). Only one study, to date, is consistent with the notion of
higher-level differences in brain structure or function as an explanation of
gender differences in odor-naming ability. Lehrner
(Lehrner, 1993), found that
women were better able to recognize suprathreshold odors at time intervals
varying from 30 min to 3 weeks in a recognition memory paradigm. Interestingly
and consistent with a biological explanation of this type, familiarity (as
indexed by the ability of participants to label odors) was not the source of
the gender difference in recognition memory.
The experiments reported here examined whether there are gender differences in the acquisition, retention or proneness to interference of unfamiliar odor—name pairs. This approach should reduce the impact of prior experience as the relatively unfamiliar odors and names should not have been encountered before. However, if experiential effects manifest motivationally, then female participants should acquire the odor—name associations faster than men. The effect of differences in odor sensitivity should also manifest in terms of acquisition rate. If female participants are better at forming a representation of the olfactory stimulus and thus better able to discriminate the odors, this should also speed acquisition. Differences in retention were established by retesting participants following a 1 week interval. In experiment 2 a further phase explored the effects of retroactive interference. Gender differences in retention and proneness to interference would be more suggestive of some form of memory difference, such as encoding, forgetting, retrieval or consolidation effects. This would likely favor a biologically based explanation, centered upon structural or functional differences in the brains of men and women.
|
Experiment 1 |
|---|
|
TopAbstractIntroductionExperiment 1Experiment 2General discussionReferences |
|---|
Experiment 1 used a set of seven odors, each of which was paired with a novel Swahili word (equated for novelty and pronounceability in a pilot study). Participants received two training trials with each odor, in which the name was provided by the experimenter. They then received 10 further blocks of trials, with each block containing seven trials and each trial consisting of one of the seven odors (sampled without replacement). On each trial participants sniffed the odor, were asked to name it and were then given feedback. Each block of trials was timed. After completing the final block of trials, participants received an association test, in which each of the seven odors was presented, with a list of the seven odor names. One week later participants returned and received two more blocks of trials, identical to those described above (i.e. give name, receive feedback), followed by a final association test.
Materials and methods
Participants
Thirty-six Macquarie University psychology undergraduates participated for
course credit. There were 16 males (mean age = 20.6 years; range 18-35 years)
and 20 females (mean age = 22.9 years; range 17-47 years).
Odors and names
Based on results from pilot data, seven unfamiliar odors were selected
(values in parentheses indicate the amount of odorant placed on a cotton wool
ball in each blue opaque plastic squeezy bottle): 1-octonal (0.70 g); mandarin
aldehyde (4.0 g); patchouli (0.06 g); bornyl acetate (0.32 g); acetyl methyl
carbinol (0.06 g); phenyl acetylene (0.16 g); and methyl anthranilate (0.40
g).
The odor names were Swahili words selected by pilot testing for novelty and pronounceability. The words were: kabali, watu, juma, kesho, vitabu, siri and mabaya.
Procedure
Experiment 1 was conducted over two sessions, separated by a 1 week
interval (see Table 1). On the
first session (day 1) participants received two learning blocks, ten test
blocks and an association test.
|
The first learning block consisted of seven trials, with a new
odor—name pair presented on each trial. Assignment of odors to names was
random. Presentation order across blocks was also random. Each learning block
trial started with the participant smelling the target odor for 
2 s (it
was self paced). This involved placing the tip of the plastic squeezy bottle
2-3 cm below the nose and sniffing while the bottle was squeezed. After the
odor bottle was returned to the experimenter, she read out the odor's name,
then spelt it out, while participants copied it down on to their response
sheet. This ensured they fully attended to the name. The participant then
sniffed the odor a second time and made three ratings, each on a seven-point
category scale. First, whether they had ever smelled the odor before (anchors;
1 = no, 4 = unsure, 7 = yes). Second, how familiar the odor was (anchors; 1 =
unfamiliar, 7 = very familiar). Third, how strong the odor smelled (anchors; 1
= no smell, 7 = very intense). This procedure was repeated for the six
remaining odor—name pairs. There was then a 2 min interval which was
followed by an identical second learning block.
On completion of the second learning block, participants received another 2 min interval before the first test block commenced. Each test block also consisted of seven trials.
On each test block trial, a participant sniffed one of the seven odors and was asked to provide its Swahili name and rate how confident they were in their judgment. Confidence ratings are not reported here as they revealed little of interest in either experiment. If the participant correctly named the odor, they were told `That's correct'; if they were incorrect, the appropriate Swahili name was provided (`No, that's X'). After repeating this process for the remaining six odors, there was an interval of 1.5 min, followed by the next test block. Each test block was timed and participants moved at their own pace (see Figure 2 for test block length). This pattern was then repeated until all ten test blocks had been completed.
|
Following a 2 min interval after the final test block, participants received an association test. This was composed of seven trials. On each trial one of the seven odors was presented along with a list of the seven Swahili names. Participants task was to select the correct name. No feedback was provided in this test. The test was timed.
The second session was completed 1 week later in the same room with the same experimenter (R.A.D.). The session commenced with two test blocks identical in form to those from the first session. This was followed by an association test, again identical to the one from the first session.
Analysis
As in both experiments reported in this paper, all data met the necessary
assumptions for parametric testing.
Results
Odor characteristics
The ratings of familiarity and intensity obtained on the first learning
block, collapsed across the different odors, did not significantly differ
between males and females (independent t-test). Collapsed across
odor, the mean ratings were: ever smelled before, male = 4.1/7 (SE = 0.2),
female = 4.2/7 (SE = 0.3); familiarity, male = 4.1/7 (SE = 0.2), female 4.0/7
(SE = 0.3); and intensity, male = 4.7/7 (SE = 0.1), female = 4.7/7 (SE =
0.2).
Correct naming during acquisition
Figure 1 depicts mean
percent correct naming, across test blocks, on sessions 1 (day 1) and 2 (day
8). The data from session 1 were analyzed by a two-way repeated-measures
ANOVA, with one between factor (Gender) and one within factor (Test block).
The analysis revealed a main effect of Test block [F(9,34) = 28.69,
P < 0.001], indicating, as can be seen in
Figure 1, that correct naming
increased across test blocks. Importantly, there were no effects involving
Gender.
|
Naming speed during acquisition
Figure 2 illustrates mean
speed in seconds, at which men and women produced names for the seven odors
(en masse), in each test block, on sessions 1 (day 1) and 2 (day 8).
The data from session 1 were also analyzed by a two-way repeated-measures
ANOVA, with one between factor (Gender) and one within factor (Test block).
Although production times significantly decreased over Test blocks
[F(9,34) = 13.59, P < 0.001], there were no effects
involving Gender.
Association test
Mean percent correct responses did not significantly differ between men and
women (independent t-test; males = 46%; females = 57%). There was no
significant difference in test length (s) between males and females
either.
Retention of names
Figure 1 also shows percent
correct naming following the 1 week interval. It is readily apparent that men
performed worse following the interval than women. This was confirmed by a
two-way ANOVA, with one repeated-measure Test block (block 10 versus block 11)
and one between factor Gender. The analysis revealed a significant interaction
between Gender and Test block [F(1,34) = 12.44, P <
0.001], indicating poorer retention in men following the interval. There was
also a main effect of Test block [F(1,34) = 38.35, P <
0.001], but no main effect of Gender.
Recovery
By the second test block of day 8 (see
Figure 1) there was no longer a
significant difference in naming performance between men and women
(independent t-test).
Naming speed and retention interval
These data are illustrated in Figure
2. There was no significant change in naming speed between test
block 10 and 11, nor any difference by gender (using ANOVA). In addition,
there was no significant gender differences in naming speed on the final test
block 12 (independent t-test).
Final association test
Mean percent correct responses did not significantly differ between men and
women (independent t-test; males = 50%; females = 51%). There was no
significant difference in test length (s) between males and females
either.
Discussion
Experiment 1 revealed that men and women can acquire novel odor—name associations at the same rate. This suggests that the ability to learn odor names is not impaired by either differences in sensitivity between males and females or by overt differences in motivation. The key difference to emerge was that following the 1 week retention interval. Here, male participants were significantly worse at correctly naming the target odors than female participants. This retention effect appears analogous to typical laboratory naming tasks, as they employ familiar odors (as were the test odors here at that point) which may not have been recently smelled or named. However, this analogy is complicated by the intrusion of the association test between test block 10 and 11. This is because the association test could have inadvertently functioned as an interference task. Some evidence for this can be found in the fact that male performance on the first association test was worse, though not significantly so, than female performance. Consequently a second experiment was conducted to determine: (i) if this retention interval effect could be repeated under circumstances where there was no intervening association task; and (ii) where the effects of interference could be more explicitly examined.
|
Experiment 2 |
|---|
|
TopAbstractIntroductionExperiment 1Experiment 2General discussionReferences |
|---|
Experiment 2 employed the same training procedure as experiment 1, except that the association test at the end of the first session was dropped. As in experiment 1, a second session took place 1 week after the completion of the first. This second session started with two further test blocks, as had experiment 1. This was followed by an interference phase in which participants learned new odor—name pairs, i.e. the same odors and names were used, but were randomly reassigned for each participant. This was followed by two further test blocks to determine how well participants had acquired these new odor—name associations. Finally, this was followed by two further test blocks, in which participants were asked to provide the original odor names, to test for retention of the old material following interference.
Materials and methods
Participants
Thirty-two Macquarie University psychology undergraduates participated for
course credit. There were 16 males (mean age = 22.6 years; range 18-34 years)
and 16 females (mean age = 22.6 years; range 18-45). No participant had taken
part in experiment 1.
Materials
These were identical to experiment 1.
Procedure
The design of experiment 2 is illustrated in
Table 2. Session 1 was
identical in all respects to experiment 1, except there was no association
test. One week later, session 2 commenced with two test blocks, again
identical to experiment 1. Following a 2 min interval, participants then
received two new learning blocks. Although the design of these new learning
blocks was the same as for experiment 1, the odor—name pairs were
re-randomized, with the caveat that no pair should remain the same (i.e. if
patchouli had been named vitabu in the original learning blocks, it
could not be named vitabu in the new learning blocks). After another
2 min interval, two further test blocks followed. These were identical in
design to experiment 1, except that participants were asked to recall the
new names they had just learned. Finally, after a further 2 min
interval, participants completed two more test blocks, again identical to
experiment 1, except now participants were asked to retrieve the names that
they first learned — the old names.
|
Results
Odor characteristics
The ratings of familiarity and intensity obtained on the first learning
block, collapsed across the different odors, did not significantly differ
between males and females (independent t-test). Collapsed across
odor, the mean ratings were: ever smelled before, male = 4.0/7 (SE = 0.2),
female = 4.5/7 (SE = 0.3); familiarity, male = 3.7/7 (SE = 0.2), female 3.9/7
(SE = 0.2); and intensity, male = 4.1/7 (SE = 0.2), female = 4.4/7 (SE =
0.3).
Naming during acquisition
Figure 3 shows mean percent
correct naming, across test blocks, on sessions 1 (day 1) and 2 (day 8). The
data from session 1 were analyzed by two-way repeated-measures ANOVA, with one
between factor (Gender) and one within factor (Test block). The analysis
revealed a main effect of Test block [F(9,30) = 31.62, P
< 0.001], indicating, as can be seen in
Figure 3, that performance
increased across test blocks on day 1. Importantly, there were no effects
involving Gender.
|
Speed of naming during acquisition
Figure 4 illustrates mean
speed in seconds, at which men and women produced names for the seven odors
(en masse) in each test block, on sessions 1 (day 1) and 2 (day 8).
The session 1 data were also analyzed by a two-way repeated-measures ANOVA,
with one between factor (Gender) and one within factor (Test block). Although
production times significantly decreased over Test blocks [F(9,30) =
5.79, P < 0.001], there were no effects involving Gender.
|
Retention of names
Figure 3 also shows percent
correct naming following the 1 week interval. It is again apparent, as in
Experiment 1, that male participants performed worse following the interval
(i.e. block 11) than women. This was confirmed by a two-way ANOVA, with one
repeated-measure Test block (block 10 versus 11) and one between factor
Gender. This analysis revealed a significant interaction between Gender and
Test block [F(1,30) = 15.55, P < 0.001], indicating
poorer retention in men following the interval. There was also a main effect
of Test block [F(1,30) = 26.14, P < 0.001] and a main
effect of Gender [F(1,30) = 4.19, P < 0.05].
Recovery
By the second test block of day 8 (see
Figure 1) there was still a
significant difference in naming performance between men and women
[t(30) = 2.05, P < 0.05], unlike in experiment 1 where
this difference had dissipated.
Speed of naming and retention interval
These data are illustrated in Figure
4. There were no significant changes in naming speed between Test
block 10 and 11, nor any difference by Gender (using ANOVA). In addition,
there was no significant difference by gender in naming speed on Test block 12
(independent t-test).
Acquisition of new odor—name associations
Figure 3 illustrates the
difference between men and women for the acquisition of the new
odor—name associations on session 2 (day 8; blocks 13 and 14). A two-way
ANOVA (Gender and Test blocks) revealed a significant main effect of gender
[F(1,30) = 4.54, P < 0.05], with women performing
slightly better than men. There were no other significant effects.
Speed of naming for new odor—name associations
Figure 4 illustrates the
mean time in seconds for participants to produce the new names on test blocks
13 and 14. There were no significant differences between males and females nor
between blocks (using ANOVA).
Effects of interference
The effects of interference can be gauged by comparing test block 12 (last
test prior to interference) with test block 15 (first test post-interference).
A two-way ANOVA (Gender and Test block 12 versus 15) revealed a main effect of
Gender [F(1,30) = 5.30, P < 0.05]. Crucially, there was
no interaction of Gender and Test block, which would have implied a
differential effect of interference on males and females. Rather these results
indicate that the general difference observed following the retention interval
had still not dissipated. Finally, a comparison of male and female performance
on the last test block 16, revealed no significant difference (independent
t-test).
Effects of interference on naming speed
These data are illustrated in Figure
4. Using the same ANOVA strategy as above revealed no effects of
Test block (12 versus 15) or Gender. Males and females did not significantly
differ on the last block either (independent t-test).
Discussion
The pattern of results from this experiment were largely identical to experiment 1. The principal gender difference was again the decrease in naming performance which occurred following the 1 week retention interval in men. In this experiment, however, the effect could not be attributed to the intervening association test, strongly implying that the retention difference is caused by the passage of time. One further gender difference was also obtained. Males were poorer at learning new odor—name associations. The other purpose of experiment 2 was to determine if susceptibility to interference, an important cause of forgetting, might account for the retention interval effects. Interference had no differential impact on male/female performance, suggesting that it is unlikely to account for this effect. In summary, the key findings from this experiment were that the effect of retention interval was observed under conditions where it could not be attributed to any intervening experimental manipulation and that interference was an unlikely explanation of this effect.
|
General discussion |
|---|
|
TopAbstractIntroductionExperiment 1Experiment 2General discussionReferences |
|---|
The experiments reported here examined gender differences in the acquisition and retention of novel Swahili names for a set of unfamiliar odors. In both experiments 1 and 2, male and female participants learned odor—name associations at the same rate and did not differ on a test where names were provided (experiment 1). However, following a 1 week retention interval, naming was significantly poorer in male participants in both experiments, although this difference diminished somewhat over subsequent blocks of trials. In experiment 2, further blocks of trials established that women were slightly better at acquiring a new set of odor—name associations, but that these new associations did not have any differential impact on recalling the original set of names. That is, men appeared no more susceptible to interference than women.
The drop in performance observed in men following the 1 week interval is,
as noted before, analogous to the finding of gender differences in naming when
more familiar odors are employed. This analogy stems from the fact that in
most studies of naming, varying delays of hours, days or weeks must intervene
between having last smelled or named a particular odor and the test where it
is presented. The effect of delay in the current experiment was not trivial,
as the effect size d was 1.0 for experiment 1 and 1.3 for experiment
2, both classified by Cohen (Cohen,
1969) as large (i.e. d > 0.8). (d was
calculated by subtracting the block 11 score from the block 10 score,
separately for males and females. The female difference score was then
subtracted from the male difference score and the product divided by the
largest of the SDs from the means used in its calculation.) To put this in
some perspective, these effect sizes are considerably bigger than the average
effect sizes for the most well established gender difference, that of spatial
ability [largest average d = 0.7;
(Voyer et al.,
1995)]. Incidentally, the effect size observed in our experiments
is also larger than those seen in typical odor naming studies. Using available
data, we estimated ds of 0.4
(Doty et al., 1985;
Engen, 1987) and 0.7 (Segal
et al., 1993) for three typical naming studies. There may be at least
two reasons for the larger effect size here. First, the retention interval
between last exposure to an odor and its name was fixed here, while in typical
naming studies it would vary between individuals. Second, the use of
overlearned highly familiar odors for testing, like those found in the UPSIT,
might mask or reduce gender differences in naming.
In the Introduction, two main types of explanation were suggested to
account for gender differences in naming. First, those involving differences
in experience between men and women (Cain,
1980) and second, those based upon a biological difference in
verbal or olfactory ability (Engen,
1987; Larsson,
1997). The results from our experiments bear on these accounts in
a number of ways. First, as noted above, the apparent parallel between the
findings from this study and previous demonstrations of better odor naming in
women suggest to us that experiential factors probably have little direct role
in accounting for gender difference in naming. There are two reasons for this.
(i) The stimuli used here were equally unfamiliar to males and females, as
were their arbitrary names. This would obviate against any benefit of practice
effects in female participants, that might readily contaminate studies using
more familiar odors. (ii) There was no evidence consistent with male
participants being less motivated to learn than female participants. Male
participants were as quick in supplying names as females, no differences
emerged across acquisition trials and all participants returned for the second
test session a week later. In addition, the finding mentioned in the
introduction, that gender differences in odor naming were largely independent
of culture (Doty et al.,
1985), may also imply that experiential factors play a more
limited role.
A second implication of these findings concerns the observation that women
are more sensitive to olfactory stimuli than men, when menstruation is a
random variable, as in these experiments
(Koelega and Koster, 1974).
One place that a sensory difference should have emerged would have been during
acquisition, especially if female participants were better able to
discriminate the target odors than men. This should have had the effect of
speeding up acquisition in female participants. However, as is clearly
apparent in Figures
1,2,3,4,
there is little evidence to suggest that differences in olfactory sensitivity
or discriminability played any significant role in acquisition.
A third implication of these findings concerns some more specific type of
olfactory perceptual difference. Although there is no evidence here that
encoding differed between males and females during acquisition, the fact that
performance was so much worse following the retention interval suggests a
memory related effect. Such an effect could manifest in at least three ways.
First, it could be taken to indicate a specific retrieval deficit in male
participants. However, this explanation is problematic, as there are currently
no theories of retrieval that offer a distinction between retrieval following
a relatively short retention period, such as over the test blocks within a
session and retrieval following much longer periods of time, such as that
following the one week interval (Blake
et al., 2000). Consequently, there is no clear
theoretical basis to assert retrieval deficits as a cause for this effect.
A second possibility is that males are more prone to interference over the
retention interval, thus subsequently affecting their ability to remember the
odor—name associations on test. Two pieces of evidence suggest this is
unlikely. First, there is no reason to believe that participants would
encounter any of the target words or odors during the week interval,
providing little opportunity for interference to occur. Second, in experiment
2, retroactive interference had no differential gender effect. In fact, the
relative absence of an interference effect reconfirms Lawless and Engen's
finding that paired associate learning with odors is typically resistant to
such a manipulation (Lawless and Engen,
1977). Thus it is unlikely that differential proneness to
interference could readily account for these findings.
A third possibility concerns memory consolidation effects. These involve
the processing of recently acquired memory traces, by, for example,
facilitating long-term potentiation
(Stickgold et al.,
2001). Consolidation ultimately results in the incorporation of
the trace into long-term memory (Sutherland and McNaughton, 2000), with these
effects typically occurring across time periods of hours or days following the
original learning episode (Stickgold,
1998). If consolidation processes were less effective in men, this
could manifest in a number of ways. First, it could reduce odor recognition
accuracy as a result of poorer consolidation of the original odor trace; this
would reduce naming accuracy as male participants would be more likely to
confuse odors. Second, it could result in weaker odor-name associations, so
that even if a male participant were able to recognize the target odor, the
association to the name would be inoperative.
Although the experiments reported here were not designed to differentiate
these two possibilities, the interference phase in session 2 of experiment 2
has an interesting bearing upon this issue. If male participants did have
weaker odor—name associations by virtue of poorer consolidation, this
should have produced less interference with new odor—name
learning and consequently better acquisition. Yet male participants were worse
at acquiring new name—odor associations in experiment 2 than female
participants (see Figure 3;
blocks 13 and 14). However, this finding, is consistent with men having
difficulties in discriminating the target odors, resulting from poorer
consolidation of the odor trace. Further evidence in favor of this hypothesis
can be found in the study by Lehrner
(Lehrner, 1993) cited in the
Introduction. He observed that male participants were poorer at recognizing
odors over various delays, an effect attributed primarily to memorial rather
than to conceptual factors (associations to names, response bias, etc.). This
would suggest that male participants may have been less effective at
consolidating the olfactory memory trace into long-term memory.
In conclusion, the experiments reported here demonstrate that male participants are poorer at naming odors following a 1 week interval between acquisition and testing, even though initial rates of name acquisition did not differ between genders. Although this finding suggests that gender differences in odor naming may result from poorer memory consolidation in male participants, the current experiments can not fully delineate whether this resulted from failure to consolidate the olfactory trace or the odor—name association (or a combination thereof). More intriguingly, the results observed here could represent a more general impoverishment of memory consolidation in men. Nonetheless, the effects reported here still demonstrate one of the largest gender differences currently identified in the psychological literature.
|
Acknowledgments |
|---|
Thanks to Harry Stevenson and Roger Turner of Dragoco and Quest International, for kindly supplying some of the odorants used here.
|
References |
|---|
|
TopAbstractIntroductionExperiment 1Experiment 2General discussionReferences |
|---|
Blake, R.V., Wroe, S.J., Breen, E.K. and McCarthy, R.A. (2000) Accelerated forgetting in patients with epilepsy, evidence for an impairment in memory consolidation.Brain , 123,472 -483.
Cain, W.S. (1980) Chemosensation and cognition. In H. van der Starre (ed.), Olfaction and Taste VII. IRL Press, London, pp.347 -357.
Cain, W.S. (1982) Odor identification by
males and females: predictions vs performance. Chem.
Senses, 7,129
-142.
Cohen, J. (1969). Statistical power analysis for the behavioral sciences. Academic Press, New York.
Deems, D.A., Doty, R.L., Settle, G., Moore-Gillon, V., Shaman,
P., Mester, A.F., Kimmelman, C.P., Brightman, V.J. and Snow, J.B.
(1991) Smell and taste disorders, a study of 750 patients
from the University of Pennsylvania Smell and Taste Center. Arch
Otolaryngol Head Neck Surg., 117,519
-528.
Doty, R.L. (2001) Olfaction.Annu. Rev. Psychol. , 52,423 -452.[Web of Science][Medline]
Doty, R.L., Shaman, P., Applebaum, S.L., Giberson, R., Sikorski,
L. and Rosenberg, L. (1984) Smell identification
ability: changes with age. Science,226
, 1441-1443.
Doty, R.L., Applebaum, S., Zusho, H. and Settle, G. (1985) Sex differences in odor identification ability: a cross-cultural analysis. Neuropsychology,23 , 667-672.[Web of Science][Medline]
Doty, R.L., Snyder, P.J., Huggins, G.R. and Lowry, L.D. (1981) Endocrine, cardiovascular and psychological correlates of olfactory sensitivity changes during the human menstrual cycle. J. Comp. Physiol. Psychol.,95 , 45-60.[Web of Science][Medline]
Engen, T. (1987) Remembering odors and their names. Am. Scient., 75,497 -503.
Feingold, A. (1992) Sex differences in variability in intellectual abilities: a new look at an old controversy.Rev. Ed. Res. , 62,61 -84.
Gilbert, A.N. and Wysocki, C.J. (1987) The smell survey. National Geographic,172 , 514-525.
Hedges, L.V. and Nowell, A. (1995)
Sex differences in mental test scores, variability and numbers of
high-scoring individuals. Science,269
, 41-45.
Herz, R.S. and Cahill, E.D. (1997) Differential use of sensory information in sexual behavior as a function of gender. Hum. Nat., 8,275 -286.[Web of Science]
Hyde, J.S. and Linn, M.C. (1988) Gender differences in verbal ability: a meta-analysis.Psychol. Bull. , 104,53 -69.[Web of Science]
Hyde, J.S. and McKinley, N.M. (1997) Gender differences in cognition: results from meta-analyses. In Caplan, P.J., Crawford, M., Hyde, J.S. and Richardson, J.T.E. (eds),Gender Differences in Human Cognition . Oxford University Press, Oxford, pp. 30-51.
Koelega, H.S. and Koster, E.P. (1974) Some experiments on sex differences in odor perception. Ann. NY Acad. Sci., 237,234 -246.[Web of Science][Medline]
Larsson, M. (1997) Semantic factors in
episodic recognition of common odors in early and late adulthood: a
review. Chem. Senses, 22,623
-633.
Lawless, H. and Engen, T. (1977) Associations to odors: Interference, mnemonics and verbal labeling.J. Exp. Psychol. Hum. Learn. Mem. , 3,52 -59.
Lehrner, J.P. (1993) Gender differences in
long-term odor recognition memory: verbal versus sensory influences and the
consistency of label use. Chem. Senses,18
, 17-26.
Nopoulos, P.C. and Andreasen, N.C. (1999) Gender differences in neuroimaging findings. In Leibenluft, E. (ed.), Gender Differences in Mood and Anxiety Disorders. American Psychiatric Press, Washington, DC, pp.1 -24.
Schneider, R.A. and Wolf, S. (1955)
Olfactory perception thresholds for citral utilizing a new type of
olfactorium. J. Appl. Physiol., 8,337
-342.
Schleidt, M. (1980) Personal odor and nonverbal communication. Ethol. Sociobiol.,1 , 225-231.
Schleidt, M., Hold, B. and Attili, G. (1981) A cross-cultural study on the attitude towards personal odors. J. Chem. Ecol., 7,19 -31.[Web of Science]
Segal, N.L., Topolski, T.D., Wilson, S.M., Brown, K.W. and Araki, L. (1995) Twin analysis of odor identification and perception. Physiol. Behav.,57 , 605-609.[Medline]
Ship. J.A., Pearson, J.D., Cruise, L.J., Brant, L.J. and Metter, E.J. (1996) Longitudinal changes in smell identification. J. Gerontol., 51A,m86 -m91.[Web of Science]
Stickgold, R. (1998) Sleep: off-line memory reprocessing. Trends Cognit. Sci.,2 , 484-492.[Web of Science]
Stickgold, R., Hobson, J.A., Fosse, R. and Fosse, M.
(2001) Sleep, learning and dreams: off-line memory
reprocessing. Science, 294,1052
-1057.
Sutherland, G.R. and McNaughton, B. (2001) Memory trace reactivation in hippocampal and neocortical neuronal ensembles. Curr. Opin. Neurobiol.,10 , 180-186.[Web of Science]
Swaab, D.F., Chung, W.C.J., Kruijer, F.P.M., Hofman, M.A. and Ishunina, T. (2001) Structural and functional sex differences in the human hypothalamus. Horm. Behav.,40 , 93-98.[Medline]
Thakur, G.P., Shahi, S.P. and Kumar, A. (1981) Social memory task in relation to sex, age, and rural-urban background. Psychol. Rep.,49 , 948.[Web of Science]
Venstrom, D. and Amoore, J.E. (1968) Olfactory thresholds in relation to age, sex or smoking. J. Food Sci., 33,264 -265.[Web of Science]
Voyer, D., Voyer, S. and Bryden, M.P. (1995) Magnitude of sex differences in spatial abilities: A meta-analysis and consideration of critical variables. Psychol. Bull., 117,250 -270.[Web of Science][Medline]
Wallace, P. (1977) Individual discrimination of humans by odors. Physiol. Behav.,19 , 577-579.[Medline]
Witryol, S.L. and Kaess, W.A. (1957) Sex differences in social memory tasks. J. Abnorm. Soc. Psychol., 54,343 -346.[Web of Science]
Zadra, A.L., Nielsen, T.A. and Donderi, D.C. (1998) Prevalence of auditory, olfactory, and gustatory experiences in home dreams. Percept. Mot. Skills,87 , 819-826.[Web of Science][Medline]
Accepted July 15, 2002
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Larsson, L.-G. Nilsson, J. K. Olofsson, and S. Nordin Demographic and Cognitive Predictors of Cued Odor Identification: Evidence from a Population-based Study Chem Senses, July 1, 2004; 29(6): 547 - 554. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. S. Choudhury, P. Moberg, and R. L. Doty Influences of Age and Sex on a Microencapsulated Odor Memory Test Chem Senses, November 1, 2003; 28(9): 799 - 805. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||