How well can young adults and children discriminate between odors?

doi:10.21203/rs.3.rs-3827810/v1

Download PDF

Research Article

How well can young adults and children discriminate between odors?

https://doi.org/10.21203/rs.3.rs-3827810/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 18 Jun, 2024

Read the published version in Psychological Research →

You are reading this latest preprint version

It is widely believed that people can distinguish between many odors although there is limited empirical evidence. Odor discrimination tasks are employed much less often than other measures of olfaction, but, interestingly, performance is typically ~ 75% correct. This less-than-perfect performance is rarely highlighted, although it suggests that people may not be as good at discriminating odors as is commonly believed. Odor discrimination is understudied in children, and although available evidence suggests that it improves with age, children perform better when the task is simpler. In the present study we explored odor discrimination in children and young adults with a relatively simple same-different task using common and uncommon odors. We found that children perform as well as adults, but that overall performance was less than perfect and depended on the odors to be discriminated. We found evidence that ability to discriminate between odors improves as the difference in pleasantness of the odors increases. In a second experiment, we tested this directly by exploring whether odors that differ in pleasantness and edibility, two dimensions that appear to be important in olfactory perception, are easier to discriminate than odors that are the same on those dimensions. We found further evidence that odors that differ in pleasantness are easier to discriminate.

children

development

pleasantness

common odors

uncommon odors

odor discrimination

It seems self-evident to the lay person and, perhaps, to many olfactory scientists that people can easily discriminate between a large number of odors. Who can’t tell the difference between the scent of a flower and that of an onion, for example? Our anecdotal experience suggests that we can easily make such discriminations. Of course, trusting our everyday experience may be misleading because we rarely test our discriminatory power in the absence of contextual cues and input from other sensory systems. Moreover, people are notoriously poor at self-assessment of their sense of smell (Landis et al. 2003; Lötsch & Hummel, 2019).

One can find reference to people’s ability to discriminate between odors in standard undergraduate textbooks on sensation and perception, such as “…a healthy person can discriminate thousands of odors” (Wolfe et al., 2013, p. 415) and more recently “A healthy person can discriminate–tell the difference between–a huge number of odors” (Wolfe et al., 2022, p. 473). Such statements seem warranted given statements such as “Reports in the literature and, indeed, our own everyday experience suggests that we are capable of making fine discriminations” (Laska & Hudson, 1992, p. 411), and “Humans can discriminate at least 1 trillion olfactory stimuli” (Bushdid et al., 2014, p. 1). However, some textbook authors make more modest claims, such as the following from Kandel et al. (2021):

Although the discriminatory ability of humans is somewhat limited compared with that of many other animals, odor chemists estimate that the human olfactory system may be capable of detecting more than 10,000 different volatile chemicals. Perfumers who are highly trained to discriminate odorants can distinguish as many as 5,000 different types of odorants… (p.862)

It is interesting to note that these authors mention that humans can detect many odors but are somewhat more circumspect about their discriminatory ability.

What is the evidence on people’s ability to discriminate between odors?

So, how well can people discriminate – tell the difference between – odors? Data on odor quality discrimination is relatively sparse compared to that of odor threshold/sensitivity and odor identification. However, one good source of odor discrimination data is studies that use the standard Sniffin’ Sticks test battery (e.g., Hummel et al., 1997). This test battery includes a 3-alternative forced-choice (3AFC), oddball or odd-one-out odor discrimination task in which the test odor is presented with two other, identical, non-target odors. The task of the participant is to pick the “target” (oddball) out of the set of three stimuli. Data from the original Sniffin’ Sticks study (Hummel et al., 1997) suggest that, for the set of odors selected, odor discrimination performance was only about 75-80% correct, well below what one might expect if people really can discriminate between thousands of odors. The absolute level of performance on the Sniffin’ Sticks test battery has received relatively little attention, although it is consistent among studies that have used Sniffin’ Sticks (e.g., Boesveldt et al., 2008; Hedner et al., 2010; Hummel et al., 2007a; Sabiniewicz et al., 2021; Sorokowska et al., 2014; Seo et al., 2011; Wolfensberger et al., 2000; Yang et al., 2010) and others that have used different methods (e.g, DeWijk & Cain, 1994; Doty et al., 2018; Hulshoff Pol et al., 2000; Laing & Francis 1989; Laska, 2004; Potter & Butters 1980; Rabin, 1988; Weierstall & Pause, 2012). Some studies have reported odor discrimination better than 90%, particularly for familiar odors (Savic & Berglund, 2000).

The experiments described in the current study start with this finding – that odor discrimination performance isn’t perfect – and we explore variables that affect it. It is worth pointing out that we are highlighting the imperfection in odor discrimination performance whereas other studies have explored whether discrimination performance is better than chance. For example, data from Bushdid et al. (2014), who reported that people can discriminate among a very large number of odors, report performance well below 100%, although significantly above chance.

Same-Different, 2AFC and 3AFC Tasks

The absolute level of odor discrimination performance (and whether it is perfect or better than chance) should take into consideration the task of participant. It appears that most tests of odor discrimination in the literature use a 3AFC task, as described above. In such a task, chance is 33.3%. Perhaps this method is preferred because all trials are equivalent, whereas in 2-alternative forced-choice (2AFC) or same-different tasks, half of the trials include pairs of “same” stimuli and the other half “different” stimuli. The task of the participant is to indicate whether the stimuli are the same or different and in such tasks chance is 50%. Moreover, the “same” trials don’t provide information about the ability to discriminate between odors per se, but rather only about whether the same odor can be confused with itself.

Chance performance level is important if one wants to compare across tasks and when statements are made about overall performance. One study in the literature compared performance on the standard Sniffin’ Sticks test battery with 16 trials in either a 2AFC (same-different) or 3AFC (triangle/oddball) task with children between 5 and 17 years of age (Gellrich et al. 2017). They reported that, in addition to performance improving with age, it was significantly higher on a 2AFC than a 3AFC odor discrimination task. However, the data reported do not appear to have been corrected for guessing. When we applied a correction for guessing to the data presented in Table 1 of Gellrich et al. (2017), we found that performance in both tasks (averaged across age groups) was ~62% correct (9.9/16). However, in the youngest age group (5–7 year olds), performance may have been higher in the 2AFC task (M_corrected=7.8/16 or 49%) compared to the 3AFC task (M_corrected=6.25/16 or 39%). It is important to note that reports of odor discrimination performance – that is, the 75–80% value cited above – cannot be interpreted without knowledge of the task and certainly cannot be directly compared across task without correcting for guessing. For the purposes of comparing across tasks, 80% correct on the standard 3AFC task is 70% when corrected for guessing.

In the experiments described here, one of which involved testing young children (Experiment 1a), we have opted to employ the cognitively simpler 2AFC or more accurately – a same-different – task, as described below. In our task, chance is 50% although we have been more interested in determining how far performance is from being perfect.

Odor discrimination or odor memory over a short delay?

Odor discrimination tasks are essentially odor memory tasks with very short delay since it is impossible to smell two things simultaneously. Or perhaps better stated, odor memory tasks are odor discrimination tasks with a delay and odor memory performance, even over very short delays, is not perfect. To that point, Zucco et al. (2014) have demonstrated that standard odor discrimination tasks depend on short-term memory and that allowing for unlimited sampling of odors can improve performance, particularly in the young and very old.

In one test of odor memory, Choudhury et al. (2003) demonstrated that the ability to match an odor to a sample (given 4 choices) did not vary as a function of delay, although it was impacted by both sex and age. The task employed by Choudhury and colleagues (previously referred to as a memory task, but now referred to as a memory/discrimination task) uses the same four common odors (peanut, rose, banana and mint) on every trial – one of the odors serves as the target odor and rest as foils/distractors. Interestingly, the best performance on this task (among the youngest female participants) was ~10.2 out of 12 trials correct. Corrected for guessing, that comes out to ~80% correct, roughly consistent with many of the studies of odor discrimination cited above.

Wenzel et al. (2021) tested memory over very short delays using an odor recognition task with three conceptual distinct categories - “sweet” spices (clove, nutmeg and ginger), “savory” spices (oregano, rosemary, basil), and three uncommon odors (diethyl malonate, vigoflor and 5-methylheptan-3-one oxime). In the forced-choice condition in which the two test odors were presented immediately after the target and the task was to say whether one of the test odors (the target and a foil) was the target or neither of them was the target, performance was significantly higher when the foil odor came from a different category than the target odor. In their no delay condition (effectively a discrimination task), overall performance was about 74% correct. Corrected for guessing (given that it was a 3AFC task), performance was 61%. Even for the odors that were discriminated the best, performance only reached 70% correct. In both of the studies described above, performance was not perfect, demonstrating that people have difficulty remembering them over very short delays or, perhaps, discriminating between them.

Odor discrimination performance depends upon odor pairs

Most studies do not report odor discrimination performance by odor pair, although those that do report that performance depends upon the stimuli or more specifically, the pairs of odors that are being discriminated. For example, Boesveldt et al. (2008) found that discrimination ranged from 44% correct for (-) and (+) carvone to 87.3% correct for the discrimination between 2-phenyl ethanol and isoamyl acetate. Note that chance was 33.3% in this study. Likewise, Potter and Butters (1980) observed very high performance (d’=4) for the discrimination of n-Butyric acid and eugenol, but lower performance (d’=1.8) for the discrimination of pineapple and apple, using a same-different task. In a triangle task that compared pairs of mixtures made from 6 different odorants (coumarin, capric acid, eugenol, geraniol, phenylethyl alcohol and vanillin), Weierstall and Pause (2012) found that performance ranged from 38% to 71%. Using enantiomers, Laska (2004) showed that some pairs of odors are discriminated better than others. Whereas performance was over 80% for limonene, performance was not better than chance (33.3%) for limonene oxide.

Laska and Hudson (1992) examined people’s ability to discriminate between pairs of odors that varied in the number of shared components in a same-different task. They found a positive correlation between the absolute number of shared components and discriminability of odor mixtures and that participants mistakenly indicated that “same” stimuli were “different” on over 50% of trials, and were no better than 80% correct on the trials in which the stimuli were different. Although these authors stated that people are capable of “fine discriminations” as noted above, they also indicated that performance in their study was “considerably poorer than expected” (p. 403).

Finally, Zhang et al. (2017) demonstrated that the development of odor discrimination in children depended upon odor. Odor discrimination improved in young children for some but not all odors and some odors were discriminated equally well by their youngest (3.5 years) and oldest (5.5 years) children.

In sum, it appears that people are less than perfect at discriminating between odors and that discrimination depends, not surprisingly, on the particular odors that are being compared. The data seem inconsistent with our intuition that we can easily tell the difference between odors.

The goal of the present study was to explore how well people can discriminate between common odors, as well as some uncommon odors, in the absence of contextual cues on a simple same–different task. We hypothesized that people would perform significantly worse than 100% correct on an odor discrimination task, but better than chance (50%). We also explored whether several variables, such as age, sex and specific pairs of odors would affect odor discrimination performance. We expected that females would out-perform males and that adults would out-perform children. We hypothesized that odor discrimination would depend upon odor pair and we explored the effect of naming, pleasantness, familiarity, intensity and edibility on odor discrimination.

Experiment 1a

In this experiment, we compared odor discrimination performance between children (6-12 years old) and young adults. Whereas many studies have reported that sense of smell develops throughout childhood, the tasks used to assess performance (e.g., odor identification and detection) often require a level of knowledge and cognition/attention that disadvantage children (for a review see Cameron 2018). Cameron (2018) suggested that developing an odor discrimination task, which would be less cognitively demanding, to test olfaction in children could be informative in understanding odor perception in children.

Tests of odor discrimination in children are relatively rare, but typically report an improvement with age. For example, Stevenson et al. (2007) tested 6– and 11–year old’s ability to discriminate odors using an oddball (3AFC) task and found that performance improved with age, particularly when odors were less familiar. Richman and colleagues (1995) developed a match-to-sample discrimination task with odors familiar to children. Testing was reliable for children 5 years of age and older and they found performance improved with age. They also found that performance was less variable on odor discrimination compared to odor identification. A study by Gellrich et al. (2017), described above, reported an effect of age on odor discrimination, and also better performance in the simpler, 2AFC task (but note discussion of correction for guessing, above). Zhang et al. (2017) found that improvement in odor discrimination in very young children depended upon degree of molecular structural similarity.

We opted to use a same-different task to explore children’s ability to discriminate between odors with which children would be familiar. We wanted the task to be as simple as possible so that performance of young children would reflect their ability to smell, not their ability to understand or complete the task (see also Hummel et al., 2007b; Richman et al., 1995; Zhang et al., 2017), nor be affected by working memory load (Zucco et al., 2014). We also wanted to examine performance as a function of the pairs of odors to be discriminated. We used common odors with which we expected children to be familiar and uncommon odors (see Cameron et al., 2016; Sulmont et al., 2002; Wenzel et al., 2021) with which no participants would be familiar.

Thus, the primary motivating questions for our first experiment were: How well can children discriminate between odors using a simple same-different odor discrimination task? And, is their discrimination ability really worse than that of young adults? A secondary aim of this experiment was to examine whether odor discrimination depended upon the pairs of odors being compared. We also explored whether children and adults differed in their ratings of odor pleasantness and whether odor discrimination performance depended upon the difference in odor pleasantness rating of pairs of odors.

Method

Participants

Participants were 25 children (15 female and 10 male), whose ages ranged from 6-12 years (M = 8.2) and 22 young adults (17 female and 5 male) whose ages ranged from 18-21 years (M = 19.3). The children were recruited from an after-school program or by word of mouth and were compensated with a small school supply. The young adults were recruited from a subject pool of psychology students or by word of mouth at Carthage College and were compensated with 0.5 research credits. Five children and three young adults reported either having seasonal allergies (six), and/or experiencing allergies (three) and/or cold symptoms (two). An independent samples t-test revealed no significant difference (t(45)=.09, p =.93) between the mean discrimination performance of this subset of 8 participants (M=70.3%, SD=15.6) and that of the rest of the sample (M=70.8%, SD=12.7). Thus, all participants’ data were included in the analyses. Two of the children reported that English was not their native language, however they had no issues with understanding the task. The procedure was approved by the Carthage College Institutional Review Board, complies with the Declaration of Helsinki for Medical Research involving Human Subjects, and informed consent was obtained from each participant prior to testing. These data were collected prior to the Covid-19 pandemic.

Materials

Eight odorants were selected from a set of odorants used in a previous study from our lab (Cameron et al. 2016). Two of the odors were uncommon and comprised one pair (vigoflor and diethyl malonate). The other six odorants were common and easy to name and hence were expected to be relatively easy to discriminate. They were placed in pairs based on their conceptual or categorical similarity (orange/lemon, chocolate/cinnamon, lavender/lilac). Liquids were dropped from a 1ml disposable pipette onto 1 x 4 cm strips of filter paper and set into 30ml amber-colored glass jars and solids were placed into jars (see Table 1 for details of stimuli; odors were judged to be isointense by lab members in a pilot study). The jars were capped and the sides covered in aluminum foil to obscure their contents. To ensure fresh stimuli, jars of orange and lemon were remade daily. The other common odors were remade when their scent began to fade. Uncommon odors were not remade because their scent did not noticeably fade during the course of testing.

Table 1

The 8 Odors (4 pairs) and their Quantities Employed in Experiment 1a

Odorant	Quantity
Orange extract	0.3 mL (6 drops)
Lemon extract	0.15 mL (3 drops)
Chocolate chips	2 morsels
Ground cinnamon	0.1 g
Lavender extract	0.05 mL (1 drop)
Lilac warming oil	0.05 mL (1 drop)
Vigoflor	0.05 mL (1 drop)
Diethyl malonate	0.05 mL (1 drop)

Note. This table was created in Microsoft Word Version 16.78.3

All statistical tests were conducted using jamovi (https://www.jamovi.org/).

Procedure

Parental consent and child assent along with a short demographic and health screening form were completed prior to testing the children. Verbal instructions were given for the discrimination task and participants were made aware of a second task that would be explained later. Participants were given the option not to smell the “uncommon” odors since their safety could not be unequivocally guaranteed. A barrier was placed between the researcher and participant in order to prevent visual discrimination of jars. Odor pairs were presented birhinally for about a second in rapid succession and the task of the participant was to indicate whether they smelled the “same” or “different”. Participants completed 16 trials. Each pair of odors was presented four times (two same odor pairs and two different odor pairs) with order of stimuli counterbalanced. Upon completion of the odor discrimination trials, participants then rated each of the eight odors on their pleasantness (1 – least pleasant to 5 – most pleasant) and they attempted to provide a name for each odor. Adult participants were asked to rate their sense of smell (1 – poor to 5 – average to 9 – exceptional). Testing sessions lasted ~15-20 minutes.

Coding of Naming Data for All Experiments

The coding of naming data was completed by two of the authors (SF and ARS). They independently coded each response and then met to discuss and come to a consensus on any differences.

Code 1 – Accuracy: This code refers to how similar the name provided was to the actual name of the odor. This was used for all common odors. A zero was awarded for no label or an incorrect label (e.g., cinnamon for apple), 0.5 was awarded for a generic label or category (e.g., fruit for apple) or another member of the category (e.g., orange for apple), and 1 was awarded for a correct label, including variants on the label (e.g., apple soap or green apple for apple). Overall agreement between coders was 97% for Experiment 1a and over 99% for Experiments 1b and 2.

Code 2 – Name Provided: This code refers to the ability of participants to provide any name for the odor. This was used for all common and uncommon odors. A dichotomous scale was used, either 0 or 1. Zero was awarded if the participant was unable to provide a name for the odor. A 1 was awarded if the participant was able to provide any at all name for the odor. There were no disagreements between coders on any items.

Results and Discussion

Overall, performance on this odor discrimination task was 70.7% (SD=13.1), which is significantly lower than 100% (t(46)=-15.4, p<.001) but greater than 50%/chance (t(46)=10.9, p=.001). We observed no sex differences in either children (Mmales=69.4%, SD=14.0, Mfemales=71.4%, SD=14.4, t(23)=.35, p=.73) or adults (Mmales=70.0%, SD=14.0, Mfemales=71.1% SD=12.1, t(20)=.17, p=.87), though it should be noted that our sample size was quite small. To our surprise, there was a significant positive correlation between self-rating of sense of smell and performance on the discrimination task (r(20)=.38, p=.04), but only if the test were one-tailed and included an outlier (a single participant who performed poorly on our task and who rated themselves a “1” on our self-rating scale).

Given the lack of sex differences and our sample size, we combined data from male and female participants in the remainder of our analyses. Surprisingly, there was no significant positive correlation between age and percent correct on the odor discrimination test (r(45)=-.025, p=.57), and see Figure 1).

Figure 2 shows performance on the odor discrimination test for each of the odor pairs for children and for adults. A two-way mixed ANOVA¹ revealed a main effect of odor pair (F(3,123) = 17.58, p<.001), meaning that some odor pairs were discriminated better than others, but no main effect of age (F(1,41)=.059, p=.81) as the children’s performance (M=70.1%, SD=14.0) was as high as the adults (M= 71.1, SD=12.5). There was no interaction between age and odor pair (F(3,123)=1.25, p=.30).

A paired-sample t-test indicated a significant difference between performance on “same” trials (M=73.7%, SD=16.4) and “different” trials (M=67.7%, SD=15.8), p=.04, suggesting a slight bias towards saying that odors are the same, which is inconsistent with Laska and Hudson’s (1992) finding.

In terms of naming odors, using the coding scheme described above, we found that adults provided names for odors 81% of the time, whereas children provided names only 73% of the time. Both children and adults provided names for the uncommon odors just over half (52%) of the time. Notwithstanding their attempts to name odors, the ability to name common odors correctly was overall quite poor (Madults=54.9%, SD=20.4 and Mchildren=40.3%, SD=19.0), though an independent samples t-test indicated that the adults were significantly better at naming odors than were children (t(45)=2.54, p=.02). For both adults and children, the best named odor was chocolate (91% and 84% for adults and children, respectively) and the worst named odor was lilac (18% and 6% for adults and children, respectively). The fact that adults were better at naming common odors than children, but were not better at discriminating between them suggests that naming of odors was not particularly helpful in discriminating them in this task. This conclusion was supported by the fact that the correlation between average percent naming of odors pairs and performance on the discrimination of those pairs, computed for each participant, was positive but not significant for adults (r(64)=.09, p=.24) and was positive but also missed significance for children (r(73)=.18, p=.07). This is consistent with the recent work of Cormiea and Fischer (2023) who found that verbal labels do not improve perceptual discriminability of odors. Experiments 1b and 2 explore this relationship with a larger number of odors.

Odor pleasantness ratings ranged from 2.2 (uncommon odors) to 4.3 (chocolate) on a 5-point scale. Interestingly, the ratings of odor pleasantness by children and adults were nearly identical. Figure 3 shows the correlation between ratings of adults and children. The closeness of the data points to the unity line demonstrates just how similar the ratings were (a similar finding was reported by Kneip et al. 1931, as cited in Engen 1982) and supports previous research that has demonstrated that some aspects of pleasantness may be innate (e.g., Steiner, 1979 and Soussignan et al., 1997, as cited in Khan et al., 2007). We explored whether odor pleasantness affected discriminability by exploring whether the difference in pleasantness ratings of pairs of odors was correlated with odor discrimination performance, for each participant. There was a statistically significant positive correlation (r(182)=.18, p=.007) suggesting that the greater the difference in pleasantness between pairs of odors, the better was odor discrimination performance. Experiments 1b and 2 explore this relationship in more detail with a larger number of odors.

The data from Experiment 1a support the hypothesis that people can discriminate between odors at a level higher than chance, but that performance is not perfect and depends upon odor pair. We observed no sex differences, and interestingly, found that children were as good as adults at this task. This suggests that children’s sense of smell may be better developed than is observed when measured with odor identification or threshold testing, which may be more attentionally demanding for the children and/or may require knowledge about odors that the children have not yet gained.

These results are consistent with some of the recent findings of Oleszkiewicz et al. (2022) who examined odor discrimination in children between 4 and 12 years of age using both enantiomers and common odors. We found, as they did, that odor discrimination performance can be as high in children of all ages. However, they found no age differences in odor discrimination when odors were the most similar, whereas we found no age differences for any odor pairs. It is possible that this is because we chose odors that were all conceptually similar.

The results of Experiment 1a demonstrated that odor discrimination clearly depends upon odor pair, but the set of odor pairs was limited. In Experiment 1b, we doubled the number of odors and tested only adults. We were primarily interested in replicating and extending the previous experiment. However, given previous findings that performance on olfactory tasks may depend on familiarity (e.g., Rabin, 1988), we also included familiarity ratings in this extended replication.

Method

The methods of Experiment 1b were identical to Experiment 1a except: (1) only adults participated, (2) there were eight additional odors (four additional odor pairs, see below) and (3) in addition to rating pleasantness and attempting to name odors, participants also rated each odor on its familiarity. Participants also rated their sense of smell (1 – poor to 5 – average to 9 – exceptional).

Participants

Participants were 35 college students and one faculty member recruited from Carthage College and compensated with 0.5 research credits or a $5 gift card. There were 19 females whose ages ranged from 18 and 30 years (M=20.8) and 17 males whose ages ranged from 19 and 37 years (M=21.9). Twelve of the participants indicated some issue that might have impacted their sense of smell: depression (10) and/or were occasional smokers (three) and/or were on medication (three). An independent samples t-test revealed no significant difference (p =.91) between the mean discrimination performance of this subset of participants (M=74.2, SD=12.7) and that of the rest of the sample (M=74.6%, SD=7.8). Thus, all participants’ data were included in the analyses.

The procedure was approved by the Carthage College Institutional Review Board, complies with the Declaration of Helsinki for Medical Research involving Human Subjects and informed consent was obtained from each participant prior to testing. These data were collected prior to the Covid-19 pandemic.

Materials

The additional stimuli used in Experiment 1b included the following odor pairs: banana extract/bubblegum, motor oil/lighter fluid, Ivory soap/Brut aftershave, and ketchup/mustard, which we expected to be very familiar to participants and hence relatively easy to discriminate. See Table 2 for quantities of the stimuli. Stimuli were presented in glass jars, as described above.

Table 2

The Additional 8 Odors (4 pairs) and their Quantities Employed in Experiment 1b

Odorant	Quantity
Banana extract	0.05 mL (1 drop)
Bubblegum	0.44 g
Motor oil	0.05 mL (1 drop)
Lighter fluid	0.05 mL (1 drop)
Ivory soap	1.0 g
BRUT aftershave	0.05 mL (1 drop)
Ketchup	1.07 g
Mustard	0.5 g

Note. This table was created in Microsoft Word Version 16.78.3

Procedure

The procedure was identical to Experiment 1a except that there was a total of 32 trials, and ratings, which included both familiarity and pleasantness, were made on a 10-point scale (1 – least pleasant/familiar and 10 – most pleasant/familiar). Testing lasted approximately 30 minutes.

Results and Discussion

Performance was similar to Experiment 1a. Overall, performance was quite low (M=74.5%, SD=9.5), which is significantly lower than 100% (t(35)=-16.1, p<.001), but higher than 50%/chance (t(35)=15.5, p<.001). We observed no sex difference (Mmales=76.8%, SD=9.5, Mfemales=72.4%, SD=9.2, t(34)=-1.43, p=.16). We also observed a non-significant negative correlation between self-rated sense of smell and discrimination performance (r(33)=-.16, p=.36).

Given that we observed no sex difference, we combined the data to explore odor discrimination performance across the eight pairs of odors. Figure 4 shows the range of performance – from 62.9% for ketchup and mustard to 95.7% for cinnamon and chocolate. A one-way repeated-measures ANOVA² revealed a main effect of odor pair (F(7, 238)=11.5, p<.001), indicating that some odor pairs were easier to discriminate than others.

A paired-sample t-test indicated a significant difference between performance on “same” trials (M=77.9%, SD=11.9) and “different” trials (M=71.1%, SD=13.8), t(35)=2.33, p=.02 suggesting a slight bias towards saying that odors are the same, which is, again, inconsistent with Laska and Hudson’s (1992) finding.

Participants provided names for odors 86% of the time, including the uncommon odors (vigoflor, 84% and diethyl malonate, 76%). The ability to name the odors correctly, however, was very poor. The overall average was 35% correct and it varied from 3% correct (lighter fluid) to 83% correct (chocolate). Notwithstanding this low performance, we explored whether the ability to name common odors impacted their discriminability. The average percent naming of common odor pairs and performance on the discrimination of those pairs, computed for each participant, was positive, but narrowly missed significance (r(250)=.10, p=.055). Again, this is consistent with Cormiea and Fischer’s (2023) finding that labeling odors does not improve their discriminability.

We explored whether familiarity affected discriminability by correlating average familiarity ratings of odor pairs with performance on the discrimination of those pairs, computed for each participant, and found a significant positive correlation (r(285)=.11, p=.03), confirming previous reports that familiarity improves odor performance (Rabin, 1988). We also explored whether odor familiarity and odor pleasantness affected discriminability by exploring whether the difference in each familiarity rating and pleasantness rating of pairs of odors was correlated with odor discrimination performance, for each participant. There was a statistically significant positive correlation in familiarity (r(286)=.17, p=.002) suggesting that the greater the difference in familiarity between pairs of odors, the better was odor discrimination performance. There was also a statistically significant positive correlation in pleasantness (r(285)=.15, p=.007) suggesting that the greater the difference in pleasantness between pairs of odors, the better was odor discrimination performance.

The data from Experiment 1b are broadly consistent with Experiment 1a and support the hypothesis that people can discriminate between odors, but that performance is not perfect and depends upon odor pair. We observed no sex differences. Odor discrimination performance was correlated with the difference in naming scores (but narrowly missed significance), familiarity ratings (both average and difference), and difference in pleasantness ratings. It is worth noting that people are not very good at naming odors and their ability to name odors consistently is also poor (Cameron et al., 2016), which suggests that using naming for odor discrimination is probably minimally useful.

We wanted to explore further the finding that pleasantness (that is, the difference in pleasantness between pairs of odors) was significantly correlated with odor discrimination performance, particularly given that pleasantness appears to be an important dimension in olfactory perception (see discussion below). In Experiment 2 we explore the possibility that the ability to discriminate between odors is largely determined by their relative pleasantness.

The results of Experiments 1a and b demonstrated that odor discrimination is not perfect, even for common odors, and depends upon odor pair. Although the odors were primarily common ones, the odor pairs were conceptually quite similar (cf. Wenzel et al., 2021), which could have reduced discriminability. Moreover, Experiments 1a and b employed a limited number of pairs of odors and this was a limitation in examining correlations between odor discrimination performance and differences between ratings of odor pairs. Notwithstanding the limitations of these experiments, we did find a significant correlation between the difference in pleasantness ratings between pairs of odors and people’s ability to discriminate between them. In Experiment 2, we wanted to test this relationship more directly.

Pleasantness/Hedonics – Discriminating between and within categories

Hedonic quality is regarded as an important dimension in odor perception and odor space (e.g., Engen, 1982; Khan et al,. 2007; Richardson & Zucco, 1989; Schiffman, 1974; Schiffman et al., 1977). As Engen (1982) wrote “The most important aspect of an odor has generally been believed to be its hedonic effect.” (p. 11), and “It is primarily the quality of odor and its hedonic meaning that dominates odor perception” (p. 172). Given its importance in human olfactory perception, we wondered whether it would be a dimension used by people to discriminate between odors.

Another dimension that is clearly relevant to processing odors is edibility, given the well-known interdependence of smell and taste. We reasoned that whether or not odors are edible could also impact their discriminability. In order to simplify our experimental design, we selected odors that were either both pleasant and edible or both unpleasant and inedible. Our research question was: Are odors discriminated better when odor pairs differ on the dimensions of pleasantness and edibility compared to when they are the same on those dimensions? We also asked participants to rate odors on both of these dimensions to explore the impact of those dimensions separately in post-hoc analyses.

In addition to our previous two hypotheses (that odor discrimination would not be perfect and would depend upon odor pair), a third hypothesis for this experiment was that odor discrimination performance would be better when odor pairs were different on the dimensions of edibility and pleasantness than when they were the same on those dimensions and that the greater the difference in pleasantness rating between two odors, the greater the discriminability.

Method

Participants

Participants were 29 undergraduate students from Carthage College, whose ages ranged from 18 to 24 years (M = 19.6). There were 9 males, 18 females and two who chose to not disclose their gender. Participants were recruited from a subject pool of psychology students or by word of mouth and were tested between November 2021 and March 2022. All participants were vaccinated against COVID-19, demonstrated no cold or flu symptoms, and had not ingested anything besides water for at least one hour prior to participation.

Six of the participants had previously tested positive for COVID-19, and four of those had experienced smell loss during infection. All reported that they had recovered from their smell lost prior to testing. Five participants reported clinical depression, three reported that they currently smoked and one participant reported having undergone nasal surgery. Eleven participants reported taking medications, but none were ones that are known to impact smell function (Schiffman, 1991). An independent-samples t-test indicated that there was no significant difference between overall performance of the 12 participants who indicated no issues with smell function (M=85.7%, SD=5.71) and the 17 participants who had at least one possible condition that might have impacted smell function (M=85.4%, SD=6.06), t(27)=0.12, p=.45, one-tailed). One participant reported that English was not their native language, but they were fluent in English. As compensation, participants received 1.0 research credits or a $10 gift card. The procedure was approved by the Carthage College Institutional Review Board, complies with the Declaration of Helsinki for Medical Research involving Human Subjects and informed consent was obtained from each participant prior to testing.

Materials

Demographic and health screening information was obtained via a Google Form. A “same/different” discrimination task was created using Scratch ‘n Sniff microencapsulated odorants (Sensonics International^TM), which were presented in the form of booklets. There were 19 unique odorants (apple, banana, bubblegum, cherry, chocolate, cinnamon, clove, grape, leather, licorice, mint, motor oil, natural gas, paint thinner, pineapple, pizza, raspberry, rubber tire, smoke). These stimuli were used to create 21 odor pairs (each pair resulted in four trials, as described below) and they were presented in pairs of booklets, labeled “A” and “B”. Each odor pair was presented in one of three blocks of 28 trials.

Procedure

Given that this experiment was conducted during the COVID-19 pandemic, some safety protocols were put into place. Participants tested themselves in a small lab room with a computer and the odor discrimination test. There was no unmasked, in-person contact. The testing room was quiet and testing lasted about an hour. Participants electronically signed a consent form and completed the demographic and health screening form, including a self-rating of sense of smell on a 9-point scale (1 – poor to 5 – average to 9 – exceptional).

Verbal instructions were given prior to the task. For the next 45 minutes, participants self-administered 3 blocks of 28 trials. On each trial, the participant scratched one odor patch (e.g., 1–1) from the booklet labeled “A” and then scratched the odor patch and smelled the odor patch (e.g., 1–1) from the booklet labeled “B”. Their task was to indicate, via a Google Form, whether the two odors were the “same” or “different”. On half of the trials the stimulus pairs were the same (e.g., raspberry vs. raspberry) and on the other half, they were different (e.g., raspberry vs. pineapple). In the “within” condition, the pairs were the same on the dimensions of pleasantness and edibility (e.g., chocolate vs. banana (both pleasant and edible) or rubber tire vs. leather (both unpleasant and inedible)). In the “between” condition, the pairs were different on the dimensions of pleasantness and edibility (e.g., banana vs. leather (pleasant and edible vs. unpleasant and inedible)).

We instructed participants to wait for 10 seconds between odor trials. However, given that they were self-administering the test, we could not guarantee the intertrial interval. Therefore, based on how long it took the participant to complete each block, the imposed duration of the break between blocks varied between three and five minutes to compensate for short intertrial intervals.

After the three blocks were completed, participants rated the 19 odors on intensity, pleasantness, familiarity, and edibility (1 – least pleasant/familiar/edible and 9 – most pleasant/familiar/edible) and they were asked to try to provide a name for each odor. For these ratings and the naming task, we reused stimuli from the odor discrimination task give limited resources.

Results & Discussion

Our first hypothesis, that odor discrimination would not be perfect, was supported as a single sample t-test revealed that overall performance (M= 86.6%, SD=5.88) was significantly lower than 100% (t(28) = -12.3, p < .001). A second single sample t-test revealed that odor discrimination performance was significantly better than 50%/chance (t(28) = 33.5, p<.001). We observed a non-significant negative correlation between self-rated sense of smell and discrimination performance (r(27)=-.04, p=.84).

Our second hypothesis, that some odor pairs would be discriminated better than others, was also supported and can be seen in Figure 5. For example, chocolate and banana, and apple and natural gas were discriminated over 95% of the time, whereas clove and cinnamon were relatively poorly discriminated (~70%) and rubber tire and leather were only marginally better (~75%). There was a main effect of odor pair observed in a one-way rm-ANOVA (F(20, 560) = 4.48, p < .001).

Our third hypothesis, that odor pairs between the categories of pleasantness and edibility would be easier to discriminate than odor pairs within the categories of pleasantness and edibility, was also supported. A paired samples t-test showed that, for trials in which targets were “different”, performance was significantly higher when odors were between categories (M=91.6%, SD=9.44) compared to when they were within categories (M=82.0%, SD=8.86), (t(28) = 4.76, p < .001).

A paired-sample t-test indicated no significant difference between performance on “same” (M=86.8%, SD=8.20) and “different” trials (M=86.3%, SD=7.95), t(28)=0.25, p=.81, which is inconsistent with Experiment 1 (a and b) and Laska and Hudson (1992).

Odor rating and naming tasks were completed in part to verify that participants perceived the odors the same way that we classified them. The mean pleasantness rating of odors coded as pleasant/edible (M=6.33, SD=1.11) was significantly higher than the mean pleasantness rating of odors coded as unpleasant/inedible (M=3.82, SD=1.19; t(28)=13.9, p<.001). All odors classified as pleasant/edible were rated as pleasant (i.e., >5 on a 9-point scale) except pizza (M=4.66, SD=2.14) and all odors classified as unpleasant/inedible were rated as unpleasant (i.e., <5) except paint thinner (M=6.03, SD=2.13). Raspberry was rated as the most pleasant (M=7.89, SD=1.50) and natural gas the least (M=1.45, SD=0.91).

The mean edibility rating of odors coded as pleasant/edible (M=5.25, SD=1.23) was significantly higher than the mean edibility rating of odors coded as unpleasant/inedible (M=2.44, SD=1.18); t(28)=12.3, p<.001). All odors coded as unpleasant/inedible were rated less than 5 and most odors coded as pleasant/edible were rated higher than 5 except for clove, pizza, bubblegum, cherry and apple. Importantly, there was little overlap in the ratings of pleasant/edible and pleasant/inedible odors - only one unpleasant/inedible odor (paint thinner, M=3.83, SD=2.71) was rated as more edible than two pleasant/edible odors (pizza (M=3.52, SD=2.60) and clove (M=3.24, SD=2.37)). Notice that there was a fair amount of variability in these edibility ratings and, moreover, rating edibility on a scale of 1-9 may be less meaningful than making a yes/no decision would have been.

The odors in this study were rated as quite intense (M=6.39, SD=1.19) and unpleasant/inedible odors (M=6.87, SD=1.14) were rated as more intense than pleasant/edible odors (M=5.91, SD=1.06; t(28)=5.78, p<.001). The odors in this study were rated as moderately familiar (M=5.67, SD=1.42) and the pleasant/edible odors were rated as more familiar (M=5.98, SD=1.31) than the unpleasant/inedible ones (M=5.36, SD=1.49; t(28)=2.76, p=.01).

Although there were significant differences in the rating of pleasant/edible and unpleasant/inedible odors in terms of edibility, intensity, and familiarity, we expected that pleasantness may be most important variable for discriminating between odors. We conducted Pearson r correlations between difference ratings for each pair of odors, for each participant, and their discrimination performance. There was no correlation between discrimination performance and difference in intensity ratings (r(601)=.00, p=.50). The correlation between discrimination performance and difference in familiarity narrowly missed significance (r(601)=.06 , p=.08). Surprisingly, there was no correlation between average familiarity of odor pairs and discrimination (r(601)=.01 , p=.46). Both the correlation between discrimination performance and edibility rating difference (r(601)=.09 , p=.03) and the correlation between discrimination performance and pleasantness rating difference (r(601)=.13 , p<.001) were significant. This indicates that the bigger the difference in edibility and/or pleasantness between odor pairs, the better they were discriminated.

The ability to name odors was very poor (~26%). Mint was relatively well identified (M=89.7%) but the correct naming of all other odors varied between 0% (no one could name paint thinner) and 58.6% (chocolate). Naming of pleasant/edible odors (M=31.9%, SD=12.3) was significantly higher than naming of unpleasant/inedible odors (M=13.2%, SD=13.6, t(28)=6.01, p<.001). The correlation between average percent naming of odors pairs and performance on the discrimination of those pairs, computed for each participant, was positive and significant (r(601)=.09, p=.03). Thus, naming odors may improve the ability to discriminate them, but given that naming of odors was so poor, it seems unlikely to be the primary method by which people discriminate between odors.

In order to test whether performance on the odor discrimination task might have been due to participants’ ability to do the task based on the appearance of the odor patches, we ran a control study. Eleven participants used the identical procedure as the odor discrimination task, but instead of indicating whether two patches of odor stimuli smelled the same or different, they indicated whether they looked the same or different. They made this judgement for all three blocks of 28 trials.

Surprisingly, participants performed more poorly on the visual discrimination task (M=64.2%, SD=.11) compared to the odor discrimination task (M=86.6, SD=.06). Moreover, there was no correlation between the visual and olfaction discrimination performance r(27) = -.115, p =.62. There was only one of the 21 odor pairs (cinnamon vs clove) for which discrimination performance was better on the visual than the olfactory task. Thus, it seems unlikely that participants were using vision to discriminate among stimuli.

In sum, we found that average odor discrimination performance in this experiment was 86.6% (better than chance but not perfect). However, that average was higher than performance in Experiments 1a and b and higher than previous literature. This is likely due to the fact that in this task we deliberately included odor pairs that came from different categories (i.e., different in edibility and pleasantness), which we found were easier comparisons to make. Finally, there was a range of discrimination performance among the various odor pairs – some odor pairs were discriminated better than others (e.g., banana and chocolate were well discriminated (95% correct), whereas clove and cinnamon were less well discriminated (70% correct).

Our rating and naming data indicate that performance on odor discrimination tasks could depend upon the difference in intensity, familiarity, ability to be named, edibility or pleasantness of pairs of odors, but we suggest that difference in pleasantness may be most important or useful in discriminating between odors.

The main goal of the current study was to examine how well young adults and children can discriminate between primarily common odors. In the experiments reported here, people were asked to say whether two odors, presented in rapid succession to reduce demands on memory, were two of the same odors or two different odors. This seems like a trivial task, particularly for young adults, and yet performance, though better than chance, was less than perfect (average ~70-86% correct). This is broadly consistent with other reports in the literature (Boesveldt et al., 2008; DeWijk & Cain, 1994; Doty et al., 2018; Hedner et al., 2010; Hummel et al., 2007a; Laing & Francis, 1989; Laska, 2004; Potter & Butters, 1980; Rabin, 1988; Sabiniewicz et al.,2021; Seo et al., 2011; Weierstall & Pause, 2012; Wolfensberger et al., 2000; Yang et al., 2010). This finding is interesting because the assumption seems to be that people are very good at discriminating between odors, but the data do not support a high level of performance. Thus, we wonder whether it is accurate to say that people can discriminate between thousands of odors when performance is significantly worse than perfect, at least in the absence of contextual cues? The main point we submit here is that people do not discriminate between odors, in the absence of contextual cues, with great fidelity. We believe this is an underappreciated and perhaps underreported finding about human olfactory perception.

We found no effect of age on performance on our odor discrimination task. Children as young as 6 years of age performed as well as college students on our task. This is not entirely consistent with some previous studies, which have found that performance improves with age (Gellrich et al., 2017; Richman et al., 1995; Stevenson et al., 2007). It is possible that the simplicity of our task, the use of mostly common odors, and the fact that the children were at least 6 years of age may account for this difference.

It is interesting to note that children were not as good as young adults at naming odors, but this did not affect their ability to discriminate between odors. Moreover, overall average ability to name odors was poor in all of our experiments. Average naming performance was positively correlated with odor discrimination performance in Experiment 2, but not in 1a and 1b. These findings suggests that this sort of odor discrimination task, which does not appear to rely on the naming of odors, is cognitively less demanding and does not require a lot of experience and thus may be a good method of testing children (see Cameron, 2018). These results are broadly in accord with a recent report by Cormiea and Fischer (2023) who found that odor discrimination was not impacted by naming/verbal labels. In a similar vein, Ninenko et al. (2023) used a task in which participants had to select a shape that they had learned to match to one of four odors presented. They found that the odors best discriminated were not the odors best named. It is worth noting that the naming results reported here are consistent with other findings in the literature from English speakers, but may not be generalizable to all populations (see, for example, Majid, 2021)

Finally, it is intriguing that the pleasantness ratings made by children were effectively identical to those of adults. Moreover, there was a correlation between difference in pleasantness ratings and ability to discriminate between odors in Experiment 1a, suggesting that pleasantness may be an important and enduring dimension in human odor perception (see Results and Discussion of Experiment 1a). This is interesting, and somewhat surprising, given the fact that ratings of odor pleasantness are quite labile and can be affected in complicated ways by intensity (Doty, 1975), familiarity (e.g., Distel et al., 1999; Sulmont et al., 2002) and experience/exposure (Cain & Johnson, 1978).

We found no sex differences in discrimination performance in this study. The lack of sex differences may be due to small and unbalanced sample sizes, although sex differences have been reported in some (Boesveldt et al., 2008; Choudhury et al., 2003; Gellrich et al., 2017; Hummel et al., 2007a) but not all (Hedner et al., 2010; Laska & Hudson, 1992; Zatorre & Jones-Gotman, 1990) tests of odor discrimination/memory over short delays. Doty and Cameron (2009) reviewed the literature on sex differences in olfaction and found that they are not always observed and are more prevalent in tasks that have a linguistic component, such as odor identification. Thus, the lack of a sex difference in the current study is perhaps not surprising.

We found that self-rating was generally not correlated with performance, confirming previous literature (e.g., Landis et al., 2003; Lötsch & Hummel, 2019). The positive correlation in Experiment 1a was driven by a single participant whose performance was very low and he rated himself as having a poor sense of smell. Although people are not good judges of their sense of smell they do, at least sometimes, notice when it is very poor or if they lose it, as in the case of Covid-19 patients. The lack of correlation between self-rating and performance on olfactory tasks in people with olfactory function in the normal range reinforces the importance of testing olfactory perception and not relying on self-report measures.

We compared performance on “same” and “different” trials in all of our experiments. Whereas Laska and Hudson (1992) reported that “same” trials were incorrect over 50% of the time and different trials were less than 80% correct, our results were mixed. Performance was higher on “same” than “different” trials in Experiments 1a and 1b, there was no significant difference in Experiment 2. More research is needed to address this issue in same-different odor discrimination tasks. It could provide some evidence for biases in participants’ judgements in these tasks.

What determines how well pairs of odors are discriminated?

Discrimination performance depended upon the specific odors to be compared – some discriminations were better than others – which is consistent with several previous studies (Boesveldt et al., 2008; Laska & Hudson, 1992; Potter & Butters, 1980; Weierstall & Pause, 2012; Zhang et al., 2017). In Experiment 1(a and b) performance ranged from ~60% to nearly 90% correct and odor pairs were conceptually similar (e.g., two citrus fruits or two flowers), which presumably made the discriminations more difficult. This is consistent with Mair et al.’s (1980) finding that discriminations are more difficult for stimuli that had been previously rated to be similar than ones rated to be dissimilar. Experiment 2 showed that odors were better discriminated when they were “between” categories (~91%) than when they were “within” those same categories (~85%; see Figure 5). The odors that were best discriminated were apple (pleasant/edible) and natural gas (unpleasant/inedible) and those that were least well discriminated were clove and chocolate (both pleasant/edible) and rubber tire and leather (both inedible/unpleasant). This is consistent with Wenzel et al. (2021) who also showed that memory over short delays was best in a forced-choice task when odors were “between” rather than “within” category. Thus, it appears that both odor category and particular odor pair are important for odor discrimination. Of course, that is not to say that there was no overlap, as some odors that were within category were well discriminated and some that were between category were less well discriminated.

Some previous research has found that familiarity improves odor discrimination. For example, Savic and Berglund (2000) found better performance on a same-different odor discrimination task for odors that were previously rated to be more familiar. Jehl et al. (1995) found that exposure to unfamiliar odors increased their familiarity and this resulted in an increase in odor discrimination ability. On the other hand, Cormiea and Fischer (2023) found that familiarity did not impact odor discrimination performance and Zhang et al. (2017) found that familiarity could not account for the developmental patterns of odor discrimination that they observed in their data. Our findings on the effect of familiarity on odor discrimination performance were mixed. In Experiment 1b we found a positive correlation between both average familiarity of odor pair and the difference in familiarity between odor pair and odor discrimination performance. This was not the case in Experiment 2 – odor familiarity was not correlated with odor discrimination. The inclusion of uncommon odors and the relatively few odors employed in Experiment 1b may have contributed to this difference. Although familiarity likely does impact performance on odor discrimination tasks, it may be less important than other odor dimensions.

Another dimension that could impact odor discrimination performance is odor intensity. Although we found that inedible/unpleasant odors were rated as more intense than edible/pleasant odors, there was no correlation between the difference in intensity ratings and how well participants discriminated between odor pairs (Experiment 2). So, although intensity could be important in discriminating between odors, it did not seem to play a role in our task.

This study, instead, has demonstrated the importance of pleasantness, or more accurately, the difference in pleasantness between odors on the ability to discriminate between odors. For some time, it has been recognized that pleasantness is an important dimension of odors in human perception (e.g., Engen, 1982; Richardson & Zucco 1989; Schiffman et al., 1977). Our findings are consistent with the statement that “Human beings tend to base similarity judgments on the pleasantness of simple chemical stimuli (Schiffman et al., 1977)” (as cited in Rabin, 1988, p. 533). Our finding that there was a significant positive correlation between the difference in odor pleasantness and ability to discriminate between odor pairs in all of our experiments supports the importance of pleasantness in human odor perception (e.g., Schiffman, 1974; Schiffman et al., 1977). Although Schiffman et al. (1977) did not find that similarity judgements could be predicted by individual’s “semantic differential rating on hedonic (good-bad) or tactile (sharp-not sharp) scales” (p. 389), we did find a correlation between individual participant’s pleasantness rating and their ability to discriminate odors. Our data are consistent with Khan et al. (2007) and Engen’s (1982) conclusion that odor perception is “dominated” by hedonics.

An important caveat to Experiment 2 is that in order to simplify our experimental design, we explicitly confounded edibility and pleasantness. We found that the difference between ratings in odors pairs in edibility and pleasantness were correlated with odor discrimination. We asked people to rate odors on a 9-point scale on edibility – it is probably more appropriate to ask for a binary judgement. Future research could investigate the importance of edibility in discriminating between odors. However, we suspect that edibility may be less useful in discrimination because it is more categorical than pleasantness. For example, while two odors can always differ in pleasantness, they could both be in the same category for edibility.

In sum, the findings reported here show that performance on an odor discrimination task, in the absence of contextual cues, is not perfect, even for young adults and depends upon odor pair. The data also support the conclusion that pleasantness, a clearly important dimension in odor perception, may play an important role in people’s ability to discriminate between odors.

Conflict of Interest

We have no conflict of interests to disclose. We have full control over all raw data and we agree to allow the journal to review our data, if requested.

Data Availability

Please note that for the general readership, the data are available upon request. We did not request permission from the IRB or our participants, some of whom were children, to share raw data publicly.

Ethical Standards

This study was approved by the Carthage College Institutional Review Board and was conducted in accordance with the ethical standards laid down by the 1964 Declaration of Helsinki and its later amendments. Consent was always obtained prior to participation.

Author Contributions

All authors contributed to the study conception and design. Material preparation and data collection were performed by Sierra Follett and Autumn Rajcevich Schwer. Initial analyses were conducted by Sierra R. Follett and Autumn S. Rajcevich Schwer, but the final analyses were conducted by Leslie Cameron. The first draft of the manuscript was written by Leslie Cameron and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Acknowledgement

We would like to thank Sensonics International^TM for providing materials for Experiment 2 and the Carthage College Office of the Provost for funding.

We thank the following students for their help in study design and data collection:

Jared Bauer, Breanna Weber and Katie Wojcik who were students in a freshman honors seminar course (Experiment 1a); Austin Kloften, Brianny Tenuta and Ashley Woodman who were students in a senior thesis course (Experiment 1b); Lizeth Lara, Christina Lendzion, Steven Mancilla, Jakub Witkowski, Anna Cabay, Jasmine Fajardo, and Sabrina Nikula who were students in a senior thesis or advanced research methods course (Experiment 2).

Boesveldt, S., Verbaan, D., Knol, D, L., van Hilten, J, J., & Berendse, H, W. (2008). Odour identification and discrimination in Dutch adults over 45 years. Rhinology, 46, 131-136.
Bushdid, C., Magnasco, M. O., Vosshall, L. B., & Keller, A. (2014). Humans can discriminate more than one trillion olfactory stimuli. Science, 343(6177), 1370-1372. https://doi.org/10.1126/science.1249168
Cain, W. S., & Johnson, F., Jr. (1978). Lability of odor pleasantness: Influence of mere exposure. Perception, 7(4), 459-465. https://doi.org/10.1068/p070459
Cameron, E. L. (2018). Olfactory perception in children. World Journal of Otorhinolaryngology-Head and Neck Surgery, 4(1), 57-66. https://doi.org/10.1016/j.wjorl.2018.02.002
Cameron, E. L., Anderson, M. R., & Møller, P. (2016). The accuracy, consistency, and speed of odor and picture naming. Chemosensory Perception, 9(2), 69-78.
https://doi.org/10.1007/s12078-016-9204-4
Choudhury, E. S., Moberg, P., & Doty, R. L. (2003). Influences of age and sex on a microencapsulated odor memory test. Chemical Senses, 28(9), 799-805.
https://doi.org/10.1093/chemse/bjg072
Cormiea, S., & Fischer, J. (2023). Odor discrimination is immune to the effects of verbal labels. Scientific Reports, 13(1), 1742. https://doi.org/10.1038/s41598-023-28134-w
DeWijk, R. A., & Cain, W. S. (1994). Odor quality: Discrimination versus free and cued identification. Perception & Psychophysics, 56(1), 12-18. https://doi.org/10.3758/BF03211686
Distel, H., Ayabe-Kanamura, S., Martínez-Gómez, M., Schicker, I., Kobaykawa, T., Saito, S., & Hudson, R. (1999). Perception of everyday odors – correlation between intensity, familiarity and strength of hedonic judgement. Chemical Senses, 24(2), 191–199. https://doi.org/10.1093/chemse/24.2.191
Doty, R. L. (1975). An examination of relationships between the pleasantness, intensity, and concentration of 10 odorous stimuli. Perception & Psychophysics, 17(5), 492-496. https://doi.org/10.3758/BF03203300
Doty, R. L., & Cameron, E. L. (2009). Sex differences and reproductive hormone influences on human odor perception. Physiology & Behavior, 97(2), 213-228. https://doi.org/10.1016/j.physbeh.2009.02.032
Doty, R. L., Tourbier, I., Neff, J. K., Silas, J., Turetsky, B., Moberg, P., Kim, T., Pluta, J., French, J., Sharan, A. D., Sperlin, M. J., Mirza, N., Risser, A., Baltuch, G., & Detre, J. A. (2018). Influences of temporal lobe epilepsy and temporal lobe resection on olfaction. Journal of Neurology, 265(7), 1654-1655. https://doi.org/10.1007/s00415-018-8891-y
Engen, T. (1982). The perception of odors. Academic Press, Inc.
Gellrich, J., Stetzler, C., Oleszkiewicz, A., Hummel, T., & Schriever, V. (2017). Olfactory threshold and odor discrimination ability in children - evaluation of a modified “Sniffin’ Sticks” test. Scientific Reports, 7(1), 1928. https://doi.org/10.1038/s41598-017-01465-1
Hedner, M., Larsson, M., Arnold, N., Zucco, G. M., & Hummel, T. (2010). Normative data for the “Sniffin’ Sticks” including tests of odor identification, odor discrimination, and olfactory thresholds: An upgrade based on a group of more than 3,000 subjects. European Archives of Otorhinolaryngology, 264, 237- 243.
https://doi.org/10.1007/s00405-006-0173-0
Hulshoff pol, H. E., Hijman, R., Baaré, W. F. C., Van Eekelen, S., & Van Ree, J. M., (2000). Odor discrimination and task duration in young and older adults. Chemical Senses, 25, 461-464. https://doi.org/10.1093/chemse/25.4.461
Hummel, T., Sekinger, B., Wolf, S. R., Pauli, E., & Kobal, G. (1997). "Sniffin' Sticks": Olfactory performance assessed by the combined testing of odor identification, odor discrimination and olfactory threshold. Chemical Senses, 22(1), 39-52. https://doi.org/10.1093/chemse/22.1.39
Hummel, T., Kobal, G., Gudziol, H., & Mackay-Sim, A. (2007a). Normative data for the “Sniffin’ Sticks” including tests of odor identification, odor discrimination, and olfactory thresholds: an upgrade based on a group of more than 3,000 subjects. European Archive of Otorhinolaryngology, 264, 237-243. https://doi.org/10.1007/s00405=006-0173-0
Hummel, T., Bensafi, M., Nikolaus, J., Knecht, M., Laing, D. G., & Schaal, B. (2007). Olfactory function in children assessed with psychophysical and electrophysiological techniques. Behavioural Brain Research, 180(2), 133-138.https://doi.org/10.1016/j.bbr.2007.02.040
Jehl, C., Royet, J. P., & Holley, A. (1995). Odor discrimination and recognition memory as a function of familiarity. Perception & Psychophysics, 57, 1002-1011. https://doi.org/10.3758/BF03205459
Khan, R. M., Luk, C. H., Flinker, A., Aggarwal, A., Lapid, H., Haddad, R., & Sobel, N. (2007). Predicting odor pleasantness from odorant structure: Pleasantness as a reflection of the physical world. Journal of Neuroscience, 27(37), 10015-11023. https://doi.org/10.1523/JNEUROSCI.1158-07.2007
Kandel, E. R., Koester, J. D., Mack, S. H., & Siegelbaum, S. A. (2021). Principles of neural science (6th ed.). McGraw-Hill.
Landis, B. N., Hummel, T., Hugentobler, M., Giger, R., & Lacroix, J. S. (2003). Ratings of overall olfactory function. Chem Senses, 28, 691-694.
https://doi.org/10.1093/chemse/bjg061
Laing, D. G., & Francis, G. W. (1989). The capacity of humans to identify odors in mixtures. Physiology & Behavior, 46(5), 809-814. https://doi.org/10.1016/0031-9384(89)90041-3
Laska, M. (2004). Olfactory discrimination ability of human subjects for enantiomers with an isopropenyl group at the chiral center. Chemical Senses, 29(2), 143-152. https://doi.org/10.1093/chemse/bjh019
Laska, M., & Hudson, R. (1992). Ability to discriminate between related odor mixtures. Chemical Senses, 17(4), 403-415. https://doi.org/10.1093/chemse/17.4.403
Lötsch, J., & Hummel, T. (2019). Clinical usefulness of self-rated olfactory performance: A data science-based assessment of 6000 patients. Chemical Senses, 44(6), 357-364. https://doi.org/10.1093/chemse/bjz029
Mair, R., Capra, C., McEntee, W. J., & Engen, T. (1980). Odor discrimination and memory in Korsakoff's psychosis. Journal of Experimental Psychology: Human Perception and Performance, 6(3), 445-458. https://doi.org/10.1037/0096-1523.6.3.445
Majid, A. (2021). Human olfaction at the intersection of language, culture and biology. Trends Cogn Sci, 25(2), 111-123. https://doi.org/10.1016/j.tics.2020.11.005
Ninenko, I., Kleeva, D. F., Bukreev, N., & Lebedev, M. A. (2023). An experimental paradigm for studying EEG correlates of olfactory discrimination. Frontiers in Human Neuroscience, 17, 1117801. https://doi.org/10.3389/fnhum.2023.1117801
Oleszkiewicz, A., Behl, O., Grahl, T., & Hummel, T. (2022). Odor discrimination in children ages 4-12 years. Chemical Senses, 47, 1-8. https://doi.org/10.1093/chemse/bjac005
Potter, H., & Butters, N. (1980). An assessment of olfactory deficits in patients with damage to prefrontal cortex. Neuropsychologia, 18, 621-628. https://doi.org/10.1016/0028- 3932(80)90101-3
Rabin, M. D. (1988). Experience facilitates olfactory quality discrimination. Perception & Psychophysics, 44(6), 532-540. https://doi.org/10.3758/BF03207487
Richardson, J. T. E., & Zucco, G. M. (1989). Cognition and olfaction: A review. Psychological Bulletin, 105(3), 352-360. https://doi.org/10.1037/0033-2909.105.3.352
Richman, R. A., Wallace, K., & Sheehe, P. R. (1995). Assessment of abbreviated odor identification task for children: A rapid screening device for schools and clinics. Acta Pædiatrica, 84, 434-437. https://doi.org/10.1111/j.1651-2227.1995.tb13666.x
Sabiniewicz, A., Wrage, F., & Hummel, T. (2021). The impact of cognitive and noncognitive factors on odor discrimination performance. Journal of Sensory Studies, 36(5), e12683. https://doi.org/10.1111/joss.12683
Savic, I., & Berglund, H. (2000). Right-nostril dominance in discrimination of unfamiliar, but not familiar, odours. Chemical Senses, 25(5), 517-523.https://doi.org/10.1093/chemse/25.5.517
Schiffman, S. S. (1974). Physicochemical correlates of olfactory quality. Science, 185(4146), 112-117. DOI:10.1126/science.185.4146.112
Schiffman, S. S. (1991). Drugs influencing taste and smell perception. In T. V. Getchell, L. M. Bartoshuk, R. L. Doty, & J. B. Snow (Eds.), Smell and taste in health and disease (pp. 845-850). Raven Press.
Schiffman, S., Robinson, D. E., & Erickson, E. P. (1977). Multidimensional scaling of odorants: Examination of psychological physicochemical dimensions. Chem Senses and Flavor, 2, 375-390. https://doi.org/10.1093/chemse/2.3.375
Seo, H., Hudziol, V., Hahner, A., & Hummel, T. (2011). Background sound modulates the performance of odor discrimination task. Experimental Brain Research, 212, 305-314. https://doi.org/10.1007/s00221-011-2729-5
Sorokowska, A., Sorokowski, P., & Hummel, T. (2014). Cross-Cultural Administration of an Odor Discrimination Test. Chemosensory Perception, 7, 85-90. https://doi.org/10.1007/s12078-014-9169-0
Stevenson, R. J., Mahmut, M., & Sundqvist, N. (2007). Age related changes in odor discrimination. Developmental Psychology, 43(1), 253-260. https://doi.org/10.1037/0012-1649.43.1.253
Sulmont, C., Issanchoi, S., & Köster, E. P. (2002). Selection of odorants for memory tests on the basis of familiarity, perceived complexity, pleasantness, similarity and identification. Chemical Senses, 27, 307-317. https://doi.org/10.1093/chemse/27.4.307
Weierstall, R., & Pause, B. M. (2012). Development of a 15-item odour discrimination test (Düsseldorf odour discrimination test). Perception, 41, 193-203. https://doi.org/10.1068/p7113
Wenzel, E. S., Cameron, E. L., Møller, P., & Köster, E. P. (2021). Novelty detection in memory for common and uncommon odors over short delays. Chemical Senses, 46, 1-9. https://doi.org/10.1093/chemse/bjab013
Wolfe, J. M., Kluender, K. R., Levi, D. M., Bartoshuk, L. M., Herz, R. S., Klatzky, R., Lederman, S. J., & Merfeld, D. M. (2013). Sensation & Perception (3rd ed.). Sinauer Associates.
Wolfe, J.M., Kluender, K.R., Levi, D.M., Bartoshuk, L.M., Herz, R.S., Klatzky, R.L., & Merfeld, D.M. (2022). Sensation and Perception (6th ed.) Sinauer Associates.
Wolfensberger, M., Schnieper, I., & Welge-Lüssen, A. (2000). Sniffin’Sticks®: A new olfactory test battery. Acta Otolaryngology, 120, 303-306. https://doi.org/10.1080/000164800750001134
Yang, L., Wei, Y., Yu, D., Zhang, J., & Liu, Y. (2010). Olfactory and gustatory function in healthy adult Chinese subjects. Otolaryngology-Head and Neck Surgery, 143, 554-560. https://doi.org/10.1016/j.otohns.2010.05.010
Zatorre, R.J., & Jones-Gotman, M. (1990). Right-nostril advantage for discrimination of odors. Perception & Psychophysics, 47(6), 526-31. https://doi.org/10.3758/BF03203105
Zhang, X., Chen, W., Li, S., & Zhou, W. (2017). Developmental fine-tuning of human olfactory discriminability. Chemical Senses, 42(8) 655-662. https://doi.org/10.1093/chemse/bjx047
Zucco, G. M., Hummel, T., Tomaiuolo, F., & Stevenson, R.J. (2014). The influence of short-term memory on standard discrimination and cued identification olfactory tasks. Journal of Neuroscience Methods, 222, 138-141. https://doi.org/10.1016/j.jneumeth.2013.11.006

One child and three adult participants opted not to smell the uncommon odors, and thus their data were not included in this analysis. The pattern of results was unchanged by the removal of these data. These data were analyzed in SPSS.
One participant opted not to smell our uncommon odors, and thus their data were not included in this analysis.

No competing interests reported.

Download PDF

Journal Publication

published 18 Jun, 2024

Read the published version in Psychological Research →

Editorial decision: Revision requested
29 Jan, 2024
Reviews received at journal
29 Jan, 2024
Reviewers agreed at journal
24 Jan, 2024
Reviewers invited by journal
24 Jan, 2024
Editor assigned by journal
18 Jan, 2024
Submission checks completed at journal
07 Jan, 2024
First submitted to journal
01 Jan, 2024

You are reading this latest preprint version

How well can young adults and children discriminate between odors?

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Experiment 1a

Experiment 1b

Experiment 2

General Discussion

Declarations

References

Footnotes

Additional Declarations

Status:

Journal Publication

Version 1