Establishing the content of gender stereotypes across development.

Gender stereotypes shape individuals' behaviors, expectations, and perceptions of others. However, little is known about the content of gender stereotypes about people of different ages (e.g., do gender stereotypes about 1-year-olds differ from those about older individuals?). In our pre-registered study, 4,598 adults rated either the typicality of characteristics (to assess descriptive stereotypes), or the desirability of characteristics (to assess prescriptive and proscriptive stereotypes) for targets who differed in gender and age. Between-subjects, we manipulated target gender (boy/man vs. girl/woman) and target age (1, 4, 7, 10, 13, 16, or 35). From this, we generated a normed list of descriptive, prescriptive, and proscriptive gender-stereotyped characteristics about people across the early developmental timespan. We make this archive, as well as our raw data, available to other researchers. We also present preliminary findings, demonstrating that some characteristics are consistently ungendered (e.g., challenges authority), others are gender-stereotypic across the early developmental timespan (e.g., males from age 1 to 35 tend to be dirty), and still others change over development (e.g., girls should be submissive, but only around age 10). Implications for gender stereotyping theory-as well as targets of gender stereotyping, across the lifespan-are discussed.

Introduction

Gender plays an important role in daily life. While beliefs about gender differ within and across cultures [1]. Within particular cultures, there are often gender stereotypes (e.g., behaviors, characteristics, or attributes) that are deemed to be more normative and/or desirable for one gender than another [1, 2]. Adults in the United States who violate gender stereotypes often experience social and/or economic penalties, commonly referred to as backlash [3-12]. For example, women who violate stereotypes by self-promoting on a job interview are less likely to be hired than identical men, while men who violate stereotypes by being self-effacing were less likely to be hired than identical women [8].

While the vast majority of existing work explores backlash against adults, recent research provided the first evidence that even gender-deviant children (in this case, preschoolers) experience backlash [13]: adults report liking gender non-conforming 3-year-olds less than their gender-conforming peers. This suggests that gender stereotypes have real-world consequences for adults and children alike. And yet, virtually nothing is known about how gender stereotypes change over development. Are stereotypic beliefs and expectations about young children the same as those about older children, adolescents, and adults? Are some traits consistently gendered across the lifespan, while others fluctuate? If they fluctuate, in what patterns? The present study investigates these questions, providing a novel assessment of how gender stereotypes change across the early developmental timespan.

Prior research demonstrates that, on average, adults in the United States believe that women should be communal (e.g., warm, supportive) and should not be dominant (e.g., aggressive, self-promoting); in contrast, men should be agentic (e.g., ambitious, independent) and should not be weak (e.g., passive, emotional) [12]. Violations of these gender stereotypes lead to backlash for adults in the U.S. (see [12] for a discussion). While the vast majority of empirical research on backlash has been conducted with participants in the United States, the limited data collected with participants in other countries including Australia [14], France [15], India [16], and China [17] reveal largely similar patterns (although more cross-cultural work is badly needed; see [18] for review).

Given the current data, it is unclear whether children in the United States, e.g., 1-year-old boys and girls, are held by adults to the same standards. Are they? If not, when in a child’s development do adults begin to apply gender stereotypes? Are there previously undocumented gender-stereotypes that apply only in childhood? More broadly, do gender stereotypes remain stable as children age, or do they fluctuate across development? In the present study, we asked adults to make judgments about the typicality and desirability of a large number of characteristics for targets of a wide variety of ages. This allowed us to characterize not only the presence or absence of a particular gender stereotype but also the developmental trajectories of these stereotypes.

In general, there are three types of gender stereotypes, each of which we measured in our study. Descriptive stereotypes involve characteristics that are thought to be typical of a particular gender [19]. For example, women in the United States are typically viewed as more self-aware and more anxious than men, while men are typically viewed as more extroverted and forgetful than women [2]. While individuals who violate descriptive stereotypes may surprise others, these individuals generally do not encounter backlash [19]. However, the same is not true for individuals who violate prescriptive and proscriptive stereotypes. Prescriptive stereotypes describe how members of a particular gender should behave. For example, women should be communal (e.g., cheerful, patient, and interested in children), while men should be agentic (e.g., athletic, ambitious, and assertive; [2, 12]). Proscriptive stereotypes are those that describe how members of a particular gender should not behave. For example, women should not be dominant (e.g., stubborn or rebellious), while men should not be weak (e.g., emotional or yielding; [2, 12]). Of importance, people who violate prescriptive and proscriptive stereotypes typically encounter social and economic penalties (i.e. backlash; [12, 20]).

A large body of work has sought to characterize the content of descriptive, prescriptive, and proscriptive stereotypes about adult men and women. For example, Social Role Theory [21, 22] posits that stereotypes about men and women stem directly from the sex-differentiated social roles traditionally occupied by men relative to women. Due to biological predispositions in early evolutionary history, men were more likely to be hunters, and women gatherers. Over time, people ascribed role-consistent traits to men and women, and these stereotypes took on prescriptive as well as descriptive components. However, because children are not yet capable of occupying these adult roles, it is unclear how, when, and why gender stereotypes should be applied to them. Relatedly, the stereotype content model proposes that stereotypes about adults cluster around two core dimensions: competence and warmth [18, 23, 24]. People who are viewed as high in warmth and low in competence (e.g., elderly people, housewives) elicit paternalistic stereotypes because they are perceived as low-status and non-competitive [24]. In contrast, targets who are viewed as low in warmth and high in competence (e.g., feminists, wealthy people) are met with stereotypes characterized by envy. Indeed, according to the stereotype content model, one reason that stereotype-violating women encounter backlash is that they are perceived as both high-status and competitive and therefore threaten the existing power structure. A potential challenge for this model—and one motivation for creating the current database of gender stereotypes about targets across the lifespan—is that children, by virtue of their relative lack of power, are economically, physically and socially powerless (relative to adults) and therefore are rarely thought of as threatening to the existing social order. Why, then should they experience backlash? More generally, it’s not obvious that competence and warmth are the appropriate domains for characterizing children. In short, neither theory nor empirical data suggest that gender stereotypes about children and adults are necessarily identical or even similar.

In fact, the recent work that has attempted to characterize adults’ gender stereotypes about children [13, 25] has found that gender stereotypes about children appear to differ—at least in some ways—from those about adults. One recent study found that gender stereotypes about appearance, toy preference, and communality may be present for toddlers but that many other types of stereotypes that apply to adults may not apply to young children [25]. For example, although adult men and elderly men are described as more intelligent than women, toddler girls and elementary-aged girls are described as more intelligent than boys [25]. This study provides an exciting and promising window into understanding how gender stereotypes differ across the lifespan; however, because it elicited ratings only for developmental categories (e.g., “adolescents (ages 12–18)”), it does not allow us to draw conclusions about developmental trajectories. Another recent study found that the gender stereotypes that apply to 3-year-old children are often meaningfully different from those that apply to adults: unlike for adults, traits that were rated as most typical for boys were rated as undesirable, and stereotypes about children were more likely to center around appearance than is typical for adults [13]. However, it is not possible to know whether some of these apparent developmental differences can be attributed to changes in gender stereotypes across the developmental timecourse, or whether they can instead be attributed to methodological differences across studies (e.g., in the stereotypes tested; in sample size; c.f. [25] which addresses some of these issues). More generally, these studies provide a promising starting point but do not provide a large dataset for future researchers to utilize and do not allow us to characterize the developmental trajectory of gender stereotypes.

Our study—which will catalogue gender stereotypes across development (e.g., ages 1–35)—is consequential for at least four reasons. First, in order for our general theories of gender stereotyping (e.g., Social Role Theory (e.g., [21, 22]) and the Stereotype Content Model [23, 24]) to be useful for understanding and predicting children’s learning about gender stereotypes, they must fit the data not only for adults but also for children. Second, in order to effectively study and predict gender backlash [13, 26], it is critical that we first understand the stereotypes that underlie backlash. Third, most theoretical approaches assume that gender stereotypes are learned; this implies that stereotypes could and should change over development, although there is very little data to speak to this. Fourth, from a practical perspective, adults interact with others (including children) throughout the early developmental timespan; parents, educators, and policy-makers would do well to understand the nature of the gender stereotypes that might be guiding their interactions. While evidence does suggest that people appear to encounter backlash for violating gender stereotypes across adulthood and childhood alike (e.g., [13]), for most ages research has done little to identify what these childhood gender stereotypes—and thus, their violations—even are. While we know that backlash exists, the nature of childhood gender stereotypes remain unclear, so we cannot yet predict the circumstances under which it is likely to occur across the lifespan.

In the present study, we measured adults’ gender stereotypes about infants (age one), children (ages 4, 7 and 10), adolescents (age 16), and adults (age 35). To do this, we measured gendered stereotypes about targets that ranged in age from 1 to 35. We present a database of normed gender stereotypes along with pre-registered findings generated by this study. This database fills a critical gap in the literature and will provide a set of developmental norms for researchers interested in gender development, gender backlash, and gender stereotypes.

Method

All materials, methods, and analyses were approved via Skidmore College’s IRB, and pre-registered (https://osf.io/7ahks).

Participants

Our target sample size pre-exclusions was 4,900, which we requested via TurkPrime [27]; this number was selected in order to ensure that we had approximately 100 participants per cell of our design. Participants were native English speakers who were aged 18+, who had at least a 97% approval rating for prior Mechanical Turk HITs, and who had between 100–10,000 HITs. In total, 5,260 participants consented. We did not collect demographic data from our participants and therefore cannot assess the extent to which they are representative of the general population of the United States. Consistent with our pre-registration, we excluded participants who failed to complete at least 80% of the questions (n = 308), and who failed any attention check (n = 354). This resulted in a final N of 4,598.

Design

The current study utilized a 2 [target gender: male, female] x 2 [rating type: pre/proscriptive, descriptive] x 7 [target age: 1, 4, 7, 10, 13, 16, 35] between-subjects design. This is in contrast to previous work that has manipulated these factors within-subjects [25]. In addition, our use of particular ages (e.g., “seven-year-old”) contrasts with that in previous work, which has elicited clusters of ratings (e.g., “elementary-school boys”; [25]).

We randomly assigned participants to conditions and to items within conditions. There were 175 characteristics and five attention checks, and participants were randomly assigned to see approximately 60% of the items. For each condition (e.g., for ratings of the desirability of characteristics for one-year-old boys), we obtained an average of 160 usable participants (min = 137, max = 187). For each characteristic (e.g., “pretty”), we obtained an average of 102 ratings per cell.

Materials and procedure

Our bank of characteristics consisted of 175 unique items from previous work [2, 12, 13, 28]. These included behaviors (e.g., wrestles), traits (e.g., pretty), preferences (e.g., loves pink), and appearance-related items (e.g., wears tutus).

Participants first viewed instructions as follows: “Today you will be answering questions about how [common or typical / desirable] you think particular traits are among [age] [boys/girls/men/women]”. For example, participants in one condition saw: “Today you will be answering questions about how [common or typical/desirable] you think particular traits are among 1-year-old boys”.

Participants then rated each characteristic on a 1–9 Likert-Style scale with 1 indicating “not at all X” (where X was either “desirable” or “common/typical”) and 9 indicating “very X”; 5 was labeled as “Neutral.” They were reminded of the instructions: “Indicate how typical/desirable it is in American society for [age, gender] to possess each of the following characteristics. [Scale = 1–9; 1 = Not at all Typical/Desirable; 9 = Very Typical/Desirable]”. Characteristics were randomly ordered and randomly sampled from the body of possible characteristics described above. After rating each characteristic, we collected participants’ ages and genders.

Results

As pre-registered (and as noted above), we excluded participants who were unlikely to be attending to the task by not answering enough questions or by incorrectly completing comprehension checks. We also excluded from analyses two items that were mistakenly included in our battery: “doesn’t wait his turn” (since it accidentally included a gendered pronoun) and “is clean” (since we also had the item “clean”).

Two primary goals of the current work were to develop a database of gender stereotypes across the developmental early timespan and to provide these data to other researchers. To this end, our data are available on our OSF page (https://osf.io/9pgkd/).

We had three additional specific goals, each of which was pre-registered: (1) to identify items that were and weren’t consistently gendered across the early developmental timespan (e.g., to find a list of gender stereotypes that characterize girls and women throughout development); (2) to identify items for which stereotypes changed over the early developmental timespan (e.g., items that are stereotypical at young ages but not at older ages); and (3) to identify items for which there were stereotypes at particular ages (e.g., to find a list of all gender stereotypes for 1-year-olds).

As pre-registered, we constructed linear models that predicted ratings for each characteristic from target gender (boy/girl), age (continuous), and their interaction. This allowed us to identify items where gender interacted with age (indicating that the presence and/or nature of the gender stereotype changed over development; these data are discussed later) and also items that were consistently gendered (i.e. there was no interaction with age, but rather a simple effect of gender).

Items that were not gender-stereotyped

We first report items for which we found no effect of target gender and no interaction of gender and age. In other words, these were the items for which—when considering the entirety of our dataset—we had no evidence, at any age, of gender stereotyping. For ratings of typicality, 29/175 (16.6%) characteristics that showed no effect of gender are depicted in Table 1. For ratings of desirability, 59/175 characteristics (33.71%) showed no effect of gender as depicted in Table 2. In other words, there is no evidence that these 29 traits constitute descriptive gender stereotypes (Table 1), or that these 59 traits constitute prescriptive or proscriptive gender stereotypes (Table 2).

Table 1

Characteristics showing no effect of gender in ratings of typicality.

Characteristics
Acts as a leader	Demanding	Materialistic
Ambitious	Determined	Rational
Analytical	Extroverted	Self-centered
Argues with parents	Hard-working	Self-sufficient
Assertive	Has a strong personality	Stubborn
Bratty	Has business sense	Typical
Brave	Independent	Uncertain
Childlike	Is a leader	Weak
Decisive	Is frequently sick	Willing to take a stand
Defends own beliefs	Loyal

Table 2

Characteristics showing no effect of gender in ratings of desirability.

Characteristics
Anxious	Good listener	Rational
Argues with parents	Has good manners	Refuses to pick up toys
Arrogant	Helpful	Ruthless
Bossy	Helps out around the housea	Satisfied with life
Bratty	Humble	Self-centered
Challenges authoritya	Interrupts others	Self-critical
Choosy	Is frequently sick	Slobbers
Cold towards others	Lazy	Spiritual
Competent	Likeable	Stingy
Complicated	Loyal	Stubborn
Controlling	Materialistica	Supportive
Cooperative	Moody	Thinks it’s funny when other kids are crying
Cynical	Nosy
Disobedient	Obedient	Uncertain
Does not use harsh language	Open minded	Unemotional
Enthusiastic	Persuasive	Waits turn
Excitable	Picky eater	Was an easy baby
Frequently has a runny nose	Polite	Well-behaved
Friendly	Prejudiced	Wholesome
Gets pushed around by other kids	Pulls other kids’ hair	Yielding

Note

a Despite the lack of overall effects of gender, characteristic showed a pairwise difference at least one age.

Stereotypes that were consistent across development

Ratings of typicality (i.e., Descriptive stereotypes)

We first considered items for which we found a simple effect of target gender (For brevity, we do not report main effects of age in the main paper, although readers may access our data in our repository (https://osf.io/9pgkd/); items described here with a main effect of gender may also have shown main effects of age; only those items for which (a) there was an age*gender interaction (discussed later) or (b) no main effect of gender are excluded from the reporting below.). Our preliminary analysis revealed 93 items for which there was a simple effect of target gender (and no interaction with age) on ratings of typicality. In other words, ratings of typicality for these items differed depending on whether the target was a boy or a girl, and this gender difference did not depend on target age. As pre-registered, we identified the items for which the Cohen’s d effect size of the comparison of ratings for boys vs. girls was larger than 0.4; this is in keeping with past work [2, 13]. This yielded 25 stereotypes about typicality that persisted across the developmental timeline; 15 of these met our pre-registered criteria for being descriptive stereotypes (effect size larger than 0.4 and a mean typicality above 6), while 10 did not (these met our effect size criteria but not the mean rating criteria and therefore were simply relatively more common for one gender than the other; we thus refer to them as “more typical” rather than “descriptive”). These are displayed in Tables 3 and 4.

Table 3

Characteristics rated as consistently more typical in boys/men than girls/women across the lifespan.

Characteristic	M (male)	M (female)	Classification for boys/men	Cohen’s d
Rowdy	6.54	5.11	Descriptive	0.74
Willing to take risks	6.65	5.79	Descriptive	0.46
Competitive	6.60	5.87	Descriptive	0.40
Handsomea	5.79	3.82	More typical	0.97
Dirtyb	5.74	3.88	More typical	0.90
Aggressive	5.43	4.30	More typical	0.58
Sometimes hits others	5.52	4.50	More typical	0.48
Has bruised knees	5.64	4.72	More typical	0.41

Note. All traits included here are ones for which the Cohen’s d effect size comparing ratings for boys/men vs. girls/women was larger than 0.4. Traits classified as descriptive demonstrated a mean typicality rating for boys/men above 6, while those classified as more typical demonstrated a mean below 6.

aCharacteristic is also a prescription for boys (see Table 5).

bCharacteristic is also a proscription for girls (see Table 5).

Table 4

Characteristics rated as consistently more typical for girls/women than boys/men across the lifespan.

Characteristic	M (male)	M (female)	Classification for girls/women	Cohen’s d
Enjoys wearing skirts and dresses	2.23	6.35	Descriptive	2.28
Loves pink	2.70	6.18	Descriptive	1.89
Pretty	3.86	6.59	Descriptive	1.43
Gentle	4.74	6.07	Descriptive	0.76
Tendera	4.86	6.05	Descriptive	0.66
Affectionate	5.52	6.51	Descriptive	0.55
Pays attention to appearances	4.82	6.14	Descriptive	0.55
Loves children	5.09	6.10	Descriptive	0.53
Sweet	5.51	6.41	Descriptive	0.52
Warm	5.48	6.34	Descriptive	0.52
Caring	5.53	6.33	Descriptive	0.48
Flatterable	5.41	6.21	Descriptive	0.40
Gracefula	3.59	5.22	More typical	0.88
Clean	4.33	5.82	More typical	0.76
Helps mom bake	3.91	5.39	More typical	0.66
Fragileb	4.54	5.52	More typical	0.46
Enjoys cooking	3.86	4.84	More typical	0.46

Note. All traits included here are ones for which the Cohen’s d effect size comparing ratings for girls/women vs. boys/men was larger than 0.4. Traits classified as descriptive demonstrated a mean typicality rating for girls/women above 6, while those classified as more typical demonstrated a mean below 6.

aCharacteristic is also a prescription for girls (see Table 5).

bCharacteristic is also a proscription for boys (see Table 5).

Table 5

Characteristics that were consistently rated as more desirable for one gender than the other.

Characteristic	M (male)	M (female)	Classification	Gender	Cohen’s d
Handsomea	6.92	4.59	Prescription	Boys/men	1.12
Likes to play with tools	6.50	4.97	Prescription	Boys/men	0.82
Loves sports	6.72	5.56	Prescription	Boys/men	0.67
Athletic	6.96	6.11	Prescription	Boys/men	0.48
Has a big appetite	5.91	4.76	More desirable	Boys/men	0.64
Loves to get dirty	5.76	4.63	More desirable	Boys/men	0.54
Fragileb	2.82	3.84	Proscription	Boys/men	0.52
Gracefulb	5.49	6.90	Prescription	Girls/women	0.77
Tenderb	5.85	6.68	Prescription	Girls/women	0.44
Soft spoken	4.78	5.55	More desirable	Girls/women	0.41
Dirtya	3.29	2.39	Proscription	Girls/women	0.45
Has unbrushed hair	4.02	3.22	Proscription	Girls/women	0.40

Note. All traits included here are ones for which the Cohen’s d effect size comparing ratings for girls/women vs. boys/men was larger than 0.4. Traits classified as prescriptive demonstrated a mean desirability rating above 6, while those classified as proscriptive demonstrated a mean below 4. Items that met our effect size criteria but that displayed means above 4 and below 6 are described as “more desirable” for a particular gender.

aCharacteristics that were also rated as descriptive/more typical for boys (see Table 3).

bCharacteristics that were also rated as descriptive/more typical for girls (see Table 4).

Ratings of desirability (i.e., Prescriptive and proscriptive stereotypes)

Our preliminary analysis also revealed 85 items that showed a simple effect of gender for ratings of desirability. In other words, these items were rated as consistently more desirable for one gender than the other, and this gender difference did not depend on target age. Of these, 12 met our threshold for effect size (Table 5). We found 4 items that were prescriptive for boys/men and 2 for girls/women; these were the characteristics for which there was an effect size of at least 0.4 and a desirability rating above 6. We also found 2 items that were more desirable for boys/men and 1 for girls/women (n = 1); these characteristics met our threshold for effect size but were neither rated as especially desirable (rating above 6) or undesirable (rating below 4); we refer to these as “more desirable” rather than “prescriptive.” Finally, we found traits that were proscriptive for girls/women (n = 2), and traits that were proscriptive for boys/men (n = 1); these characteristics met our threshold for effect size and had a mean desirability rating below 4.

In our previous work, we demonstrated that characteristics descriptive of 3-year-old boys also tended to be rated as undesirable (i.e., below the midpoint of our 9-point desirability scale), while the opposite was true for 3-year-olds girls [13]. We extend this finding in the present dataset; the mean desirability rating for the desirable characteristics in Table 5 for boys was 4.29 (i.e., undesirable), while it was 6.08 (i.e. desirable) for girls; these values differed significantly (p = .003). These data suggest that the characteristics that describe boys/men are rated as less desirable than those that describe girls/women (and, in fact, are rated as undesirable) across the lifespan. Further, we highlight that there were noticeably fewer traits viewed as consistently typical for boys/men (8) than for girls/women (17).

Stereotypes that change over development

We next consider the characteristics for which we found a significant gender by age interaction. These were the items for which the magnitude of the gender gap changed across the developmental timeline—in other words, traits that are gender stereotypic, but as a function of target age. We found 43 characteristics where gender differences in ratings of typicality interacted with age (Table 6) and 22 characteristics where gender differences in ratings of desirability interacted with age (Table 7).

Table 6

Characteristics for which gender differences in ratings of typicality interacted with age.

Cluster classification and included characteristics		Gender more typical
Childhood gender differences
	Likes superheroes*	Boys/men
	Pretend to be a soldier*	Boys/men
	Likes princesses*	Girls/women
	Likes to play with dolls*	Girls/women
	Wears Tutus*	Girls/women
Adolescent gender differences—boost
	Has a big appetite	Boys/men
	Has unbrushed hair	Boys/men
	Smelly	Boys/men
	Emotional*	Girls/women
	Anxious	Girls/women
	Cries often	Girls/women
	Melodramatic	Girls/women
	Moody	Girls/women
	Pays attention to what other people are wearing	Girls/women
	Self-critical	Girls/women
Adolescent gender differences—reduction
	Likes to play outside	Girls/women
	Wears clothes that don’t match	Girls/women
	Likes to be held*	Girls/women
	Choosy	Girls/women
	Snuggly	Girls/women
Fluctuating gender differences
	Comforts other children when they are crying	Girls/women
	Compassionate	Girls/women
	Eager to soothe hurt feelings	Girls/women
	Good listener	Girls/women
	Helps out around the house	Girls/women
	Sensitive to the needs of others	Girls/women
	Sympathetic	Girls/women
Cluster classification and included characteristics		Gender more typical
Fluctuating gender differences
	Understanding	Girls/women
Direction switches
	Is submissive*	Direction switches
	Spiritual	Direction switches
	Stingy	Direction switches
Differences emerge late
	Childlike	Boys/men
	Strong	Boys/men
Persistent gender differences
	Masculine*	Boys/men
	Likes to play with tools	Boys/men
	Loves to get dirty	Boys/men
	Feminine*	Girls/women
	Likes to wear nail polish*	Girls/women
Unclassified
	Sensitive*	N/A
	Steals toys*	Boys/men
	Intimidating	Boys/men
	Is easily frightened	Girls/women
	Messy	N/A
	Strong	N/A
	Unemotional	Boys/men

Note. Asterisks indicate that characteristic also had a pre/proscriptive interaction (see Table 7).

Table 7

Characteristics for which gender differences in ratings of desirability interacted with age.

Classification and included characteristics		Gender more desirable
Childhood gender differences
	Plays with trucks	Boys/men
	Likes superheroes*	Boys/men
	Pretend to be a soldier*	Boys/men
	Steals toys*	Boys/men
	Likes to play with dolls*	Girls/women
	Sensitive*	Girls/women
Differences emerge late
	Ambitious	Boys/men
	Adorable	Girls/women
	Flatterable	Girls/women
	Is submissive*	Girls/women
	Likes to be held*	Girls/women
Persistent gender effects
	Masculine*	Boys/men
	Enjoys wearing skirts and dresses	Girls/women
	Loves pink	Girls/women
	Pretty	Girls/women
	Feminine*	Girls/women
	Likes princesses*	Girls/women
	Likes to wear nail polish*	Girls/women
	Wears Tutus*	Girls/women
Unclassified
	Demanding	N/A
	Dominant	N/A
	Emotional*	N/A

Note. Asterisks indicate that characteristic also had a descriptive interaction (see Table 6).

Our next goal was to understand the nature of these interactions. To do this, as pre-registered, we visualized each interaction. We then exploratorily qualitatively clustered characteristics based on shared developmental patterns. To do this, the lead author clustered the visualizations (see Fig 1) based on visual similarity, and the other two authors checked the clustering. Due to the subjective and qualitative nature of this classification process, the resulting clusters should be interpreted as useful ways of digesting our otherwise exceptionally dense dataset, and as helpful jumping-off points for future research. Fig 1 defines and depicts each cluster. Every characteristic and its cluster are depicted in Tables 6 (for ratings of typicality) and 7 (ratings of desirability).

Fig 1

Observed clusters and their patterns.

Stereotypes at each age

As pre-registered, for each age, we classified each item according to whether it met the criteria for being a descriptive (mean rating of at least 6 for one gender, effect size of at least .4 for the difference in ratings for boys/men vs. girls/women), prescriptive (mean rating of at least 6 for one gender, effect size of at least .4 for the difference in ratings for boys/men vs. girls/women), or proscriptive stereotype (mean rating of less than 4 for one gender, effect size of at least .4 for the difference in ratings for boys/men vs. girls/women). We did this for every item whether or not there was a significant interaction of age and gender in the analyses above; this was because we decided, a priori, that it was important to identify individual stereotypes at each age. Of course, as with all situations with such a large number of comparisons, even though our threshold was not statistical significance, we encourage readers to be cautious in interpreting each particular effect as it is likely that some of these effects emerged by chance alone.

This process of classification had three main outcomes. First, we assessed how many characteristics were classified as gender stereotypes at each age (see Fig 2). Interestingly, stereotypes were most frequent between the ages of 7 and 16, peaking at age 10. While the rate of prescriptive stereotypes appeared approximately the same across the developmental timespan, proscriptive and descriptive stereotypes were most frequent during childhood in our dataset. In other words, for female targets, there were more stereotypes applied to 7-year-olds than to adults, to 10-year-olds than to adults, to 13-year-olds than to adults, and to 16-year-olds than to adults. This is particularly striking because much of the existing research has focused on understanding gender stereotypes only about adults.

Fig 2

Number of characteristics classified as gender stereotypes at each age by type of stereotype.

Notes. Y-axis is a count of stereotypes that met our pre-registered stereotype threshold. Black indicates prescriptive stereotypes, dark gray indicates proscriptions, and light gray indicates descriptive stereotypes. Note that some characteristics are double-counted (e.g., if an item was a descriptive and prescriptive stereotype at a particular age, it contributes to both the descriptive count and the prescriptive count). These data suggest that gender stereotypes are prevalent across the developmental timeline, and that children—not adults—may be subject to the most gender stereotypes.

Next, we explored whether there were more de-, pre-, or pro-scriptions for boys/men relative to girls/women and whether the frequencies of these stereotypes changed over the early developmental timespan (Table 9). We found that a larger proportion of descriptive stereotypes were about girls/women (65.4%) than about boys/men (34.6%; p < .0001). A larger proportion of proscriptive stereotypes were about boys/men (72.3%; p < .0001) than about girls/women (27.7%); there were no effects of age on the distribution of stereotypes in either of these cases (all p>.10). Interestingly, this gendered asymmetry did not emerge for prescriptive stereotypes, which were equally frequent for boys/men (49.6%) and girls/women (50.4%, p = .93). Again, there were no effects of age on the gendered distribution of these stereotypes. These data suggest that while girls/women may consistently be subject to a relatively higher proportion of descriptive stereotypes, boys/men are subject to a higher proportion of proscriptive stereotypes throughout the lifespan.

Table 9

Characteristics that were rated as prescriptive/proscriptive of one-year-old children.

Characteristic	M (girl)	M (boy)	Effect size	Classification	Other ages where effect was found
					4	7	10	13	16	35
Masculine	3.06	6.01	1.48	Prescriptive of Boys; Proscriptive of Girls	x	x	x	x	x	x
Handsome*	4.36	6.79	1.19	Prescriptive of Boys	x	x	x	x	x	x
Plays with trucks*	4.77	6.51	0.96	Prescriptive of Boys	x	x	x
Likes to play with tools	4.98	6.54	0.80	Prescriptive of Boys	x	x	x	x	x
Athletic	4.90	6.22	0.61	Prescriptive of Boys	x	x		x
Loves sports	5.11	6.16	0.56	Prescriptive of Boys	x	x	x	x	x	x
Likes superheroes	5.36	6.40	0.53	Prescriptive of Boys	x	x	x
Strong personality	5.66	6.51	0.44	Prescriptive of Boys
Enjoys wearing skirts and dresses	6.27	2.82	1.71	Prescriptive of Girls; Proscriptive of Boys	x	x	x	x	x	x
Feminine	6.28	2.99	1.71	Prescriptive of Girls; Proscriptive of Boys	x	x	x	x	x	x
Likes princesses*	6.21	3.80	1.20	Prescriptive of Girls; Proscriptive of Boys	x	x	x	x	x	x
Likes to play with dolls*	6.58	4.06	1.25	Prescriptive of Girls	x	x	x	x	x
Pretty*	6.83	4.61	1.08	Prescriptive of Girls	x	x	x	x	x	x
Graceful	6.76	5.51	0.69	Prescriptive of Girls	x	x	x	x	x	x
Sensitive	6.17	5.31	0.43	Prescriptive of Girls			x
Characteristic	M (girl)	M (boy)	Effect size	Classification	Other ages where effect was found
					4	7	10	13	16	35
Likes to wear nail polish	4.84	2.85	0.99	Proscriptive of Boys	x	x	x	x	x	x
Wears Tutus	5.30	3.34	0.96	Proscriptive of Boys	x	x	x	x	x
Loves pink	5.66	3.87	0.86	Proscriptive of Boys	x	x		x	x
Dirty*	2.56	3.54	0.47	Proscriptive of Girls	x	x		x	x
Challenges authority	3.12	4.07	0.41	Proscriptive of Girls

Note. “Other ages” column indicates the other target ages (4, 7, 10, 13, 16, 35) for which this characteristic was also descriptive.

*Characteristic was also a description for that age and gender (see Table 8).

The final outcome is a table of all items that meet the criteria for being de-, pre-, or pro-scriptive at each age range. These are available in our repository. To illustrate these findings for one age group, in Tables 8 and 9, we pull out all descriptive (Table 8) and prescriptive (Table 9) stereotypes for one-year-olds.

Table 8

Characteristics that were rated as descriptive of one-year-old children.

Characteristic	M (girl)	M (boy)	d	Gender	Other Ages
					4	7	10	13	16	35
Plays with trucksa	4.43	6.77	1.13	Boys	x	x	x
Handsomea	4.00	6.00	0.83	Boys						x
Dirtyd	5.28	6.21	0.42	Boys	x		x	x
Likes princessesb,c	6.36	3.63	1.32	Girls	x	x	x
Prettyb	6.84	4.98	0.90	Girls	x	x	x	x	x	x
Likes to play with dollsb	6.28	4.54	0.86	Girls	x	x
Does not use harsh language	7.65	6.42	0.48	Girls		x	x

Note. “Other ages” column indicates the other target ages (4, 7, 10, 13, 16, 35) for which this characteristic was also descriptive.

aCharacteristic was also a prescription for boys (see Table 9).

bCharacteristic was also a prescription for girls (see Table 9).

cCharacteristic was also a proscription for boys (see Table 9).

dCharacteristic was also a proscription for girls (see Table 9).

Discussion

In the present study, we measured adults’ stereotypes about male and female targets across the early developmental timespan (i.e. from infancy through early adulthood). To do this, we presented over 4,000 adults with a list of characteristics and asked them to rate either the desirability or typicality of those characteristics. Critically, participants rated the characteristics for targets that were either male or female and that were either 1, 4, 7, 10, 13, 16, or 35-years-old. This allowed us to develop the largest known normed database of gender stereotypes and to shed light on several questions about how descriptive, prescriptive, and proscriptive gender stereotypes change across the developmental timeline.

Rather than demonstrating stable stereotypic expectations for boys, girls, men, and women throughout the lifespan, our data revealed numerous developmental trends in the nature of gender stereotypes. First, items that were consistently gendered (main effects) were very rare; less than 10% of our items were consistently rated as being descriptive or pre/pro-scriptive stereotypes. Further, 29 characteristics were never descriptive of either gender, and 59 were never pre/pro-scriptive of either gender. These data suggest that theories of gender stereotypes need to take into account the fact that stereotypes are applied differently to targets of different ages—an idea that has not received significant attention in the literature thus far (to our knowledge).

The existence of a sizable subset of ungendered characteristics suggests that demand characteristics were unlikely to be responsible for our findings. In interpreting these results, it is important to note that we selected each of our 175 target characteristics from the existing literature [2, 12, 13, 28]—these were items for which we had strong reason to believe that stereotypes might emerge. Indeed, for some items that are considered relatively central to defining particular gender stereotypes, we found no effects whatsoever of gender (e.g., there were no gendered effects on ratings of desirability for items like helping, wholesome, is a leader, bossy, challenges authority, controlling, moody, friendly, good listener, competent or polite, all of which are items that previous work has suggested may be gendered). These data highlight the importance of empirically testing the presence or absence of gendered stereotypes one-at-a-time (in contrast to some previous work, which has clustered traits; see [25]).

While there was a sizeable subset of traits for which there was no evidence of gender stereotyping, the majority of traits did show some evidence of gendering (it is important to note that many of these effects, while statistically significant, did not reach our pre-registered effect size criteria and therefore are not reported in our main paper; they are available in our repository). Together, our data strongly suggest that most characteristics were stereotyped and that gender stereotypes change over development. Even given our stringent criteria, 43 characteristics showed significant age by gender interactions for ratings of typicality, and 22 characteristics showed significant age by gender interactions for ratings of desirability (We wish to note, of course, that “desirability” is a complex construct and certainly not a construct that is likely to be fully addressed via a single likert-scale question. We use this terminology in order to most closely approach the ways in which previous work (e.g., Prentice & Carranza, 2002) has discussed prescriptive and proscriptive gender stereotypes.). These data highlight the importance of taking a developmental approach to studying gender stereotypes. After all, our theories of the development of gender stereotypes will necessarily differ depending on whether a particular stereotype persists throughout the lifespan, emerges only in adulthood, peaks at the onset of puberty, or displays some other pattern.

Below, we discuss some of the more important developmental changes that we identified. Future research should further explore the nature of and mechanisms underscoring these changes. Additionally, we hope that other researchers will find our database immediately useful in informing the development of new research materials. For example, researchers interested in backlash targeting young adolescents will likely wish to manipulate the gender typicality of targets’ traits, but could not previously be certain about which traits are actually viewed as gender stereotypic for this particular age group. Thus, our database now enables the development of evidence-based stimulus materials conveying the gender (counter)stereotypicality of targets across the developmental timespan.

As an important first step, we qualitatively and exploratorily classified the several developmental patterns that emerged in our data. These classifications were driven by the shape of the data and not by any theoretical expectations about which stereotypes might fall into each category. Because these clusters were generated exploratorily and subjectively, we encourage future researchers to use these primarily to motivate future confirmatory research and to generate testable theories. Further, a visual inspection of our data suggests the possibility that some of the developmental changes in gender stereotyping may be non-linear. The present study was not designed to differentiate possible developmental trajectories and did not sample ages densely enough to effectively do so (e.g., step-functions vs. logarithmic vs. quadratic vs. linear; see [29] for review). Thus, we encourage researchers to explore our dataset, and to conduct further research specifically aimed at detecting differences in the shape of change of gender stereotyping across the lifespan. In addition, we describe some of the qualitative patterns that emerged in our dataset below.

First, we note the presence of the Gender Differences Emerge Late category of stereotypes; these can be found in Fig 1 and towards the bottom of Tables 6 and 7. These are items for which we identified gender stereotypes in adulthood but found that these stereotypes were minimal or absent for younger targets (e.g., ambitious, submissive). These items are important because they shed light on the existing adult gender literature in that some of these items are critical to existing theories of gender development.

The fact that there are any characteristics for which the direction of a significant perceived gender gap switches throughout the lifespan (i.e., the Gender Difference Switches cluster) is a particularly novel and surprising revelation. Additionally, it is noteworthy that some gender stereotypes appear to be strongest in adolescence (i.e., the Adolescent Gender Differences—Boost and Adolescent Gender Differences—Reduction clusters) or in childhood (the Childhood Gender Differences cluster), or in only one age group (e.g., Gender Differences Fluctuate). We have no theoretical account for these particular clusters at this moment and note that none of the clusters cleave neatly along existing theoretical lines.

Notably, we found 20 stereotypes that apply even to 1-year-old children. Of importance, from a purely developmental perspective, some of these characteristics could be difficult for children in this age group to display. For example, our data revealed that 1-year-old boys should be “athletic” and “love sports” and cannot “wear tutus” or “love pink.” It is unclear how a 1-year-old boy (who may not yet be walking and is unlikely to be talking) could adequately convey their athleticism and enthusiasm for sports, or their disdain for a color category they likely have no cognitive appreciation for and an article of clothing they are unlikely to have selected themselves. Similarly, 1-year-old girls should be “graceful” and “like princesses,” and should not be “dirty” or “challenge authority.” One might reasonably expect that infants and toddlers are too young to be constrained by these sorts of expectations, and that instead, adults would simply focus on whether very young children are healthy and meeting appropriate developmental milestones. Indeed, from this perspective, it is noteworthy that any characteristics emerged as prescriptive and proscriptive for 1-year-olds. Certainly, the current work makes the novel contribution of demonstrating that even infants appear to experience the effects of gender stereotyping.

Our results also clearly suggest that the stereotypes that individuals are faced with change over development. Of importance, existing theories of gender stereotypes were not created and thus, are unlikely to be able to account for these changes. For example, both backlash theory [30] and the Stereotype Content Model [23, 24] emphasize that adult men are expected to exhibit agentic, competence-related traits while women are expected to display communal, warmth-related traits. However, when examining the pre- and proscriptions, we uncovered that for young children it is not apparent that warmth and competence are the primary stereotypic dimensions relevant for classifying children. Instead, consistent with our prior work [13], young children’s pre- and proscriptions appear to be much more linked to appearance (e.g., clothing choices) and overt, developmentally-relevant behaviors (e.g., play preferences). This suggests that it will be important to expand the existing literature to think more broadly about the nature, fluctuation, and impact of gender stereotypes throughout the lifespan. We hope that future researchers will utilize our developmental data to hone their theories of and predictions about the origins and time course of gender stereotypes.

While much of the adult literature has focused on stereotypes about women, our data show several ways in which boys and men may experience negative gender stereotyping. First, we found that the stereotypes that were considered typical of boys and men were also more often rated as unfavorable (see also [13]); this was true across the developmental timespan studied. Second, we found that across the timespan studied, there were more proscriptive stereotypes for boys/men than for girls/women. These results build upon a growing body of work demonstrating that gender stereotypes can have profound consequences for men as well as women (for a discussion, see [4]). For example, men appear to encounter backlash when they violate gender stereotypes by expressing interest in female gender-typed careers [5], behave modestly on a job interview [6], or disclose their emotions [31]. Further, recent work has shown that adults’ reactions to 3-year-old boys who violate gender stereotypes may be particularly harsh relative to same-aged girls who violate gender stereotypes [13]. Taken together, these findings suggest that future work should continue to consider the ways in which gender stereotyping impacts perceptions of targets across the gender spectrum.

While the analyses reported here were pre-registered, we nevertheless consider them to be exploratory: we didn’t have strong predictions about which stereotypes would persists across development (e.g., we found that boys/men are more rowdy and competitive, girls/women are more flatterable and caring), which peak in adolescence (e.g., girls cry more than boys; boys have a bigger appetite than girls), which would show stereotype vacillations across the timespan (e.g., boys/men are only sometimes more stingy than girls/women), and which would show no substantial gender stereotypes at all (e.g., neither gender is more bratty, stubborn, materialistic, rational, weak, or independent). While it may be tempting to believe that some of the developmental patterns that we demonstrate are the result of noise in the data, we believe it is unlikely that the patterns we see are false positives. First, our sample sizes are large: each datapoint (e.g., the mean rating of typicality for intelligent for 3-year-olds boys) consists of 100 ratings. While it is possible for noise to be (erroneously) treated as signal, we believe that our high-powered design has likely revealed many provocative patterns of the development of gender stereotypes that were undetected in prior studies. Second, we pre-registered our data collection techniques and analyses and relied on measuring effect sizes (in addition to null hypothesis significance testing), reducing the likelihood that the patterns in our data emerged due to questionable research practices or because our design was overpowered. For these reasons, we hope future researchers will take seriously both the predicted and surprising developmental findings reported in our dataset.

We wish to note two major limitations of this study. First, we treated gender as a binary, when we know that gender is actually a continuum (see [32] for review). Tellingingly, none of our participants noted any concerns about our binarization of gender. While we do not believe that a binarized view of gender is the right one, we do believe that the average adult in the United States assumes it to be, and our participants were familiar with and able to discuss gender in a binary way. To this point, we also note that we only sampled adults in the United States. We have no reason to believe that these stereotypes generalize to other cultural contexts, and indeed, this calls for additional cross-cultural research that can provide culturally-specific information about the content of descriptive, prescriptive, and proscriptive gender stereotypes across the lifespan.

In sum, we provide a novel and rich resource for future researchers cataloging the content of gender stereotypes across the early developmental timespan. The current results highlight the importance of expanding current theories of gender stereotyping to include developmental perspectives. Simply put, current theories of gender stereotyping may be specific to one point in development (i.e., adulthood). While this is useful for informing our understanding of the ways in which gender stereotypes about adults impact perceptions of adults, additional work is needed to shed light on the ways in which gender stereotypes shape and constrain social perceptions and experiences across the lifespan. Our analyses suggest several fruitful specific directions for new programs of research (e.g., focusing on the impacts of stereotyping on infants and their caregivers; emphasizing research on boys; using developmental trajectories to inform conceptual accounts of stereotyping), and we hope that researchers will use our dataset as a resource to inform both their theoretical and empirical future work.

21 Jan 2021

PONE-D-20-25142

Establishing the Content of Gender Stereotypes Across the Lifespan

PLOS ONE

Dear Dr. Sullivan,

thank you for submitting your manuscript to PLOS ONE. After careful consideration by myself and two experts in the field, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

In your revision, please pay particular attention to the technical and statistical aspects of your work as mentioned in the Reviewwers' comments.  Also, statements of study primacy (such as "this is the first study..") have to be well-grounded.  So, please make sure to justify the original rationale of your work and discuss your work and findings in the context of existing literature such as Koenig (2018) , making clear the novelty of your research.

Please submit your revised manuscript not later than six months from this date as otherwise, any revision has to be considered a new submission. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Sasha

Alexander N. 'Sasha' Sokolov, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1.) Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: PONE-D-20-25142: Establishing the Content of Gender Stereotypes Across the Lifespan

I found this paper, which described gender stereotypes across childhood and adolescence, interesting and well-written. I would like to see the method and results of this work more directly compared to those of past studies that cataloged prescriptive and descriptive stereotypes – including Koenig (2018, who also assessed prescriptive and descriptive stereotypes across age categories, very similar to the current study) as well as Prentice (in terms of the results for adults). I believe this study adds information above these past studies, but it would be helpful if the method and results were compared to highlight how this study extends and builds on this past research. For example, I found it odd that the focus in the intro is on the stereotype content model for adults (pg. 11), which is about descriptive stereotypes, rather than this past work on prescriptive and descriptive stereotypes.

I applaud the researchers for preregistering the study, but I was unclear what exactly was preregistered. The first time preregistration is mentioned on page 13 says “preregistered yet exploratory findings” but it is unclear at this point in the paper what these exploratory analyses would be. It would be useful here to outline the three specific goals currently listed later on page 16, which would help guide the reader and understand what was preregistered.

How was the list of traits created? Some of the traits seems to apply better to children or to adults. What were participants to do when rating whether adults wear tutus (unlikely) or “gets pushed around by other kids” (did this say “other adults” when rating adults?), for example?

It might not be possible for space reasons, but it would be quite helpful if the graphs of the different “visualization pattern” of stereotypes over time were incorporated into the table listing the stereotypes with different patterns, as it is difficult to remember the various patterns. Perhaps one prototypical stereotype could be directly graphed as an example, rather than (what I assume are currently) hypothetical patterns. It was unclear how it was decided which characteristics fit each pattern – was this a visual inspection by the researchers (and if so, was this done individually and then the labels compared, as is done for qualitative codings?) or was there a quantitative analysis of similarity (or whether the effect was linear or quadratic, etc.)? Also, the characteristic “childlike” is listed twice in Table 6 under two different patterns.

I would also have liked more discussion of the findings in terms of, for example, the types/groups of traits that showed different patterns or results. Are there common themes (e.g., communion, agency, negative communion, negative agency, competence, physical characteristics, personality traits vs. behaviors) in the traits that show different results? This would make it a bit easier to digest the findings, as well as having the individual traits listed out for the specifics.

Reviewer #2: This manuscript reports a study of gender stereotypes across a variety of ages. This study appears to have been well-conducted overall. I have a few comments for the authors to consider as they move forward.

In several places, the manuscript refers to effects over "the lifespan" or "the developmental timespan," but 6 of the seven ages are children/adolescents, and the other one, 35, is used as a stand in for all adults. Truly examining this over the lifespan would sample more widely across adulthood and include older adult ages (e.g., 65, 85). Please reframe the description of these effects to more accurately reflect the represented ages.

Relatedly, for Figure 2, the caption says that the data suggest that children may be subjected to the most gender stereotypes, but again you have six ages for children and only one for all of adulthood! I think this is quite a stretch to say given the data.

It's unclear to me what "desirable" attributes are, as participants respond to these items. Are they saying they personally feel these attributes are desirable, or that they are seen in general as desirable? Do the authors have any validity evidence from this or other work using this wording that can tell us how best to understand how participants likely interpret "desirable" in this context? Literature on cultural vs. personal stereotypes would also be relevant to reference here.

The manuscript referred to preregistrations, which is great, but when I looked on OSF to see them I couldn't find them. This may be due to OSF's admittedly clunky navigation system (or my clunky inability to navigate it) but please do ensure that the preregistrations are made available.

The description of the analytical approach comes a little late - in the "Ratings of Typicality" section, but it looks like it was already employed to evaluate the "items that were not stereotyped" section. Please introduce the approach earlier to give context for how to interpret all reported results.

The Mturk sample is very large, which is great, but Mturk skews very White and likely other ways as well (e.g., politically liberal). Given that the stated aim is in part to create normed ratings of stereotypes, this should be explicitly acknowledged and discussed in the Discussion section.

There is some ambiguity about what analysis was performed, exactly (e.g., for "stereotypes that change over the developmental timespan," did the authors use regression to identify interactions, or only the descriptive method described in this paragraph?). Explicitly and clearly describing all analyses would be helpful. For this reason I chose "I don't know" about whether the analyses have been performed appropriately and rigorously, but with just a little more explanation, I would answer "Yes."

I particularly appreciate the focus on effect sizes. Could the authors explain their thinking behind choosing these particular cutoffs (e.g., why does a rating of 6 or higher count).

Social roles theory of gender stereotypes seems relevant here, especially with respect to its focus on agency and communion, which are close to warmth and competence. I would encourage the authors to reference this work if/as relevant.

The approach taken here (using established traits from literature on gender stereotypes) makes sense. For their future work, I would encourage the authors to consider whether a bottom-up approach could complement this work, as that could allow them to see if there are unique gender stereotypes that arise at different ages that are not well-captured by existing gender stereotypes (which could reflect only certain ages).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

12 May 2021

Detailed responses to reviewers are included in our cover letter.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

13 Jul 2021

PONE-D-20-25142R1

Establishing the Content of Gender Stereotypes Across Development

PLOS ONE

Dear Dr. Sullivan,

thank you for submitting your revised manuscript to PLOS ONE.  After careful consideration, we feel that it has merit but as it currently stands, still has to be improved to fully meet PLOS ONE’s publication criteria.  I appreciate very much the efforts of two experts in the filed who have provided their detailed and valuable feedback on your revision, and both been positive about your work.  Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process, paying particular attention to the following points below:

1) As you will see from the comments, both Reviewwers have had difficulty in accessing documents mentioned as pre-registered.  Please provide a correct link to the files.

2) Please make sure to address the technical and statistical issues raised by both Reviewers such as multiplicity corrections and statistical trend analyses.  In addition, please (a) state if data sets were tested for normality and which (parametric, nonparametric) statistical inference was used for respective sets, (b) which statistic (and directional or not) was used for each p-value, and (c) add variability measures where appropriate.

Also please respond thoroughly to the other concerns expressed by the Reviewers.

Please submit your revised manuscript within six moinths from this date as after that any revision has to be considered a new submission.  If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Thank you for choosing PLOS ONE for communicating your research.

Kind regards,

Sasha

Alexander N. 'Sasha' Sokolov, Ph.D.

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have revised the manuscript to address the reviewer’s concerns, making the paper even stronger and clearer. The data set is large, but the authors do a good job of aligning the analysis of the data with their goals and distilling the main points of the findings. I appreciate the thoroughness of their response to the previous reviews. I have only a few very minor suggestions for the authors to take into consideration as they move forward with editing, to make the paper as clear as possible.

Pg. 5-6: “because it elicited ratings only for developmental categories (e.g., “adolescents (ages 12-18)”),” This sentence is not complete.

Pg. 6: “in order for our general theories of gender stereotyping (e.g., Social Role Theory (e.g., Eagly & Wood, 2012; 2016) and the Stereotype Content Model; Fiske et al., 2002; Fiske et al., 1999) to be useful, they must fit the data not only for adults but also for children.” I’m not sure the authors want to imply these theories are not useful, as they were designed for stereotypes of adults and have created meaningful bodies of research. Certainly, it's possible they may not apply well to stereotypes of children (or it could be that the social roles of children do impact their gender stereotypes, as would be suggested by social role theory, which doesn't mean the stereotypes have to mirror adult stereotypes). So this sentence could be reworded to suggest these theories may need to be adapted or new theories created to address gender stereotypes across development, without throwing out these theories entirely. Otherwise, this seems like a strong point – to say these theories would not be useful if gender stereotypes differ across ages – that would need more elaboration and discussion that this one sentence.

Pg. 12: The list of traits that are not prescriptive for gender includes three items that reference "kids" – gets pushed around by other kids, pulls other kids’ hair, thinks it’s funny when other kids are crying. It is unclear whether these items were used verbatim for all groups (even adolescents and adults) or whether the item changed to reflect the age group that was being rated (e.g., “gets pushed around by other adults”).

Given the multiple tests of comparison, it would be useful to include a statement at the beginning of the results that indicates the criterion for labeling something as a significant main effect or interaction (was it p < .05?). This may be stated in the preregistration, but the links provided for the preregistration took me only to the data and I did not see a preregistration document there (although perhaps I did not know where to look - but I suggest checking the link to make sure it leads to the preregistration documents). I know it is the effect size that qualifies a characteristics as a stereotype or not, but it is discussed how many traits showed significant effects, so knowing this cutoff is still relevant.

Pg. 17: The authors reference “In our previous work, we demonstrated that..” but do not give a citation (likely for blinding for review). I would suggest now adding in a citation here so that readers know what work is being referenced.

Table 6: Perhaps in the “direction switches” category it could be indicated which gender was higher at the youngest ages and then which was high later (the “switch”), instead of indicating “N/A.”

Table 8 has blanks when the stereotype was not relevant to that age group, but Table 9 uses dashes (and one blank). Do these dashes also represent ages where the stereotype was not relevant? I would suggest using blanks in both tables as that seems the most intuitive (and I believe APA style suggests dashes when the data are not collected/irrelevant, which is not the case here).

Reviewer #2: I was a reviewer on the prior submission. I really like the aim of this project – to descriptively catalogue gender stereotypes at a range of ages. The current manuscript is stronger and has overall addressed my comments well. I have a few comments that I hope will help to strengthen the paper further.

In response to my original comment about the pre-registration (#17), I appreciate the clarification, but I still can’t find the files. When I go to the OSF page, at first the only options available to me are “Files” “Wiki” and “Analytics”. Then when clicking “Analytics”, “Registrations” appears. But then when I click on “Registrations”, I get a message “there was an error loading this list”. So, I still have not been able to review the registrations.

I really appreciate the authors moving away from phrases like “lifespan” and being clearer in identifying the ages examined in this work. However, the phrase “across the developmental timespan” (e.g., as used in the abstract and the introduction) still implies that the developmental timespan encompasses only childhood / up to middle adulthood, when in fact development continues throughout the lifespan. Perhaps a phrase like “early lifespan” or “early developmental timespan” or “from childhood through early adulthood” would be more accurate?

In the method: “For example, participants in one condition saw: “Today you will be

answering questions about how [common or typical/desirable] you think particular traits are among 1-year-old boys”.”

Should either “common or typical” or “desirable” be listed, as it’s an example, or did participants within this condition actually rate on both commonness and desirability (which is what this wording could suggest)?

I really like the visualization for the interactions. But, only a linear term of age is used, which could potentially omit some characteristics that would have quadratic/cubic effects for one gender but not another. If there are nonlinear patterns of interaction e.g., if girls tend to increase and then decrease on a trait, whereas for boys it remains flat, that won’t necessarily appear in the interactions as specified if the overall (linear) change for girls is steady. It looks from the visualizations that for many of those characteristics that showed a difference in the linear slope (i.e., a significant interaction), there was likely some difference in quadratic/cubic change, too – e.g., childhood gender differences, adolescent gender differences – boost, and fluctuating gender differences. These traits just happen to have been identified by the linear interactions because there is also an overall difference in slope between genders, and not because the models run could identify these kinds of patterns. This suggests that there may be characteristics for which males and females differ in their quadratic/cubic patterns but not in their overall linear term. Could the authors re-rerun the regressions and see if including a quadratic or cubic term for age would suggest additional characteristics to explore? This can be done by adding both age(centered)^2 as a linear term and as interacting with gender, and age(centered)^3 as a linear term and as interacting with gender. This could potentially identify additional characteristics that fit the current set of visualization patterns, and/or new ones.

Regarding: “Critically, these findings are unlikely to be due to biases in the characteristics tested in our study: if they were, we would not expect the symmetrical rate of descriptive stereotypes.” (p. 28).

I don’t quite follow this, can the authors provide a little more explanation of the logic here?

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

23 Sep 2021

Reviewer #1

Reviewer #1: The authors have revised the manuscript to address the reviewer’s concerns, making the paper even stronger and clearer. The data set is large, but the authors do a good job of aligning the analysis of the data with their goals and distilling the main points of the findings. I appreciate the thoroughness of their response to the previous reviews. I have only a few very minor suggestions for the authors to take into consideration as they move forward with editing, to make the paper as clear as possible.

We thank the reviewer for their kind words, and appreciate the remaining suggestions, which we have responded to below.

Pg. 5-6: “because it elicited ratings only for developmental categories (e.g., “adolescents (ages 12-18)”),” This sentence is not complete.

The sentence now reads “This study provides an exciting and promising window into understanding how gender stereotypes differ across the lifespan; however, because it elicited ratings only for developmental categories (e.g., “adolescents (ages 12-18)”), it does not allow us to draw conclusions about developmental trajectories.”

Pg. 6: “in order for our general theories of gender stereotyping (e.g., Social Role Theory (e.g., Eagly & Wood, 2012; 2016) and the Stereotype Content Model; Fiske et al., 2002; Fiske et al., 1999) to be useful, they must fit the data not only for adults but also for children.” I’m not sure the authors want to imply these theories are not useful, as they were designed for stereotypes of adults and have created meaningful bodies of research. Certainly, it's possible they may not apply well to stereotypes of children (or it could be that the social roles of children do impact their gender stereotypes, as would be suggested by social role theory, which doesn't mean the stereotypes have to mirror adult stereotypes). So this sentence could be reworded to suggest these theories may need to be adapted or new theories created to address gender stereotypes across development, without throwing out these theories entirely. Otherwise, this seems like a strong point – to say these theories would not be useful if gender stereotypes differ across ages – that would need more elaboration and discussion that this one sentence.

Thank you for pointing out this unfortunate wording. We do not wish to imply that the SCM isn’t useful. We now say: “First, in order for our general theories of gender stereotyping (e.g., Social Role Theory (e.g., Eagly & Wood, 2012; 2016) and the Stereotype Content Model; Fiske et al., 2002; Fiske et al., 1999) to be useful for understanding and predicting children’s learning about gender stereotypes, they must fit the data not only for adults but also for children”

Pg. 12: The list of traits that are not prescriptive for gender includes three items that reference "kids" – gets pushed around by other kids, pulls other kids’ hair, thinks it’s funny when other kids are crying. It is unclear whether these items were used verbatim for all groups (even adolescents and adults) or whether the item changed to reflect the age group that was being rated (e.g., “gets pushed around by other adults”).

Thank you for pointing out this ambiguity. These items were used verbatim.

Given the multiple tests of comparison, it would be useful to include a statement at the beginning of the results that indicates the criterion for labeling something as a significant main effect or interaction (was it p < .05?). This may be stated in the preregistration, but the links provided for the preregistration took me only to the data and I did not see a preregistration document there (although perhaps I did not know where to look - but I suggest checking the link to make sure it leads to the preregistration documents). I know it is the effect size that qualifies a characteristics as a stereotype or not, but it is discussed how many traits showed significant effects, so knowing this cutoff is still relevant.

We apologize again for the error in the preregistration link. Our alpha was .05, and all tests were two-tailed. Our effect-size cutoffs were also pre-registered.

Pg. 17: The authors reference “In our previous work, we demonstrated that..” but do not give a citation (likely for blinding for review). I would suggest now adding in a citation here so that readers know what work is being referenced.

Thank you. We have now added our citation.

Table 6: Perhaps in the “direction switches” category it could be indicated which gender was higher at the youngest ages and then which was high later (the “switch”), instead of indicating “N/A.”

When we attempted to implement this suggestion, we found that it looked quite clunky (e.g., “higher for girls at ages 1 and 4, and higher for boys for age 7+”). In addition, the analyses required to determine the precise nature of the “switch” are different from the analyses used in the rest of the table (the former involves pairwise comparisons within age, while the latter involves regression models over the entire dataset); we were concerned that reporting particular age information in this table might confuse readers.

However, we agree with the reviewer that “N/A” is not the ideal notation for this table. We now say “direction switches” in this column. Readers who are interested in understanding the precise nature of the switch can access our data and supplemental materials.

Table 8 has blanks when the stereotype was not relevant to that age group, but Table 9 uses dashes (and one blank). Do these dashes also represent ages where the stereotype was not relevant? I would suggest using blanks in both tables as that seems the most intuitive (and I believe APA style suggests dashes when the data are not collected/irrelevant, which is not the case here).

We have now made our formatting consistent, by changing all dashes to blanks.

Reviewer #2: I was a reviewer on the prior submission. I really like the aim of this project – to descriptively catalogue gender stereotypes at a range of ages. The current manuscript is stronger and has overall addressed my comments well. I have a few comments that I hope will help to strengthen the paper further.

Thank you! We really appreciate the time you’ve put into improving the manuscript.

In response to my original comment about the pre-registration (#17), I appreciate the clarification, but I still can’t find the files. When I go to the OSF page, at first the only options available to me are “Files” “Wiki” and “Analytics”. Then when clicking “Analytics”, “Registrations” appears. But then when I click on “Registrations”, I get a message “there was an error loading this list”. So, I still have not been able to review the registrations.

We are really sorry -- we (accidentally!) had two OSF pages for the same project with very similar titles; one had the data and the other had the preregistration. The correct preregistration link is now throughout the article, and we have moved all datasets etc… from the other OSF page to the correct one. As noted above, the correct link to the pre-registration is: https://osf.io/7ahks and the correct link the OSF page containing that pre-registration is: https://osf.io/9pgkd/

I really appreciate the authors moving away from phrases like “lifespan” and being clearer in identifying the ages examined in this work. However, the phrase “across the developmental timespan” (e.g., as used in the abstract and the introduction) still implies that the developmental timespan encompasses only childhood / up to middle adulthood, when in fact development continues throughout the lifespan. Perhaps a phrase like “early lifespan” or “early developmental timespan” or “from childhood through early adulthood” would be more accurate?

Thank you. We have changed it to “early developmental timespan” in most cases, and to the “developmental timepsan studied” when we are referring to the ages tested in our study.

In the method: “For example, participants in one condition saw: “Today you will be

answering questions about how [common or typical/desirable] you think particular traits are among 1-year-old boys”.”Should either “common or typical” or “desirable” be listed, as it’s an example, or did participants within this condition actually rate on both commonness and desirability (which is what this wording could suggest)?

It now says “...questions about how [“common or typical”][“desirable”] you think particular traits are among…”

I really like the visualization for the interactions. But, only a linear term of age is used, which could potentially omit some characteristics that would have quadratic/cubic effects for one gender but not another. If there are nonlinear patterns of interaction e.g., if girls tend to increase and then decrease on a trait, whereas for boys it remains flat, that won’t necessarily appear in the interactions as specified if the overall (linear) change for girls is steady. It looks from the visualizations that for many of those characteristics that showed a difference in the linear slope (i.e., a significant interaction), there was likely some difference in quadratic/cubic change, too – e.g., childhood gender differences, adolescent gender differences – boost, and fluctuating gender differences. These traits just happen to have been identified by the linear interactions because there is also an overall difference in slope between genders, and not because the models run could identify these kinds of patterns. This suggests that there may be characteristics for which males and females differ in their quadratic/cubic patterns but not in their overall linear term. Could the authors re-rerun the regressions and see if including a quadratic or cubic term for age would suggest additional characteristics to explore? This can be done by adding both age(centered)^2 as a linear term and as interacting with gender, and age(centered)^3 as a linear term and as interacting with gender. This could potentially identify additional characteristics that fit the current set of visualization patterns, and/or new ones.

We agree that this is a fascinating possibility, and one that is ripe for examination in future research. However, comparing models in these suggested ways would add a substantial amount of complexity and information to the paper, which is already dense. Adding to the complexity is that these analyses would not only be exploratory, but also atheoretical: while we acknowledge that there are some visible non-linearities in our data, we know of no theories that predict a particular cubic or quadratic effect of age -- or, critically, an interaction of target gender with a cubic or quadratic age function. This makes it challenging to explore this question in a targeted way -- when conducting exploratory analyses that deviate from the plan, it is best practice to do so when theoretically motivated.

More generally, our ability to make fine-grained determinations about the shape of developmental change hinges on the density of our age sampling; while our sampling of different ages across childhood allows us to make general statements about the overall trajectory, without fine-grained sampling of ages (i.e., at least of every year), any descriptions of the shape of developmental change will be impacted substantially by our sampling frequency (for an excellent description and demonstration of this in the domain of motor development, see Adolph et al., 2008).

Our goal is to provide some preliminary observations about trends in these data, and then make the dataset available for other researchers to examine in numerous different ways. As such, we have added the sentence to the General Discussion on pp. 31-32. :

“Further, a visual inspection of our data suggests the possibility that some of the developmental changes in gender stereotyping may be non-linear. The present study was not designed to differentiate possible developmental trajectories and did not sample ages densely enough to effectively do so (e.g., step-functions vs. logarithmic vs. quadratic vs. linear; see Adolph, 2008 for review). Thus, we encourage researchers to explore our dataset, and to conduct further research specifically aimed at detecting differences in the shape of change of gender stereotyping across the lifespan. In addition, we describe some of the qualitative patterns that emerged in our dataset below”

Regarding: “Critically, these findings are unlikely to be due to biases in the characteristics tested in our study: if they were, we would not expect the symmetrical rate of descriptive stereotypes.” (p. 28). I don’t quite follow this, can the authors provide a little more explanation of the logic here?

Upon revisiting that paragraph, we also are unsure what we intended with this sentence -- we have removed it from the manuscript.

Submitted filename: DevTraj_Reponse_Sept.docx

Click here for additional data file.

17 Jan 2022

Establishing the Content of Gender Stereotypes Across Development

PONE-D-20-25142R2

Dear Dr. Sullivan ,

thank you for your revised manuscript above.  We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Thank you for submitting your research to PLOS ONE.

Kind regards and stay safe and healthy in 2022,

Sasha

Alexander N. 'Sasha' Sokolov, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Rebecca Neel

15 Feb 2022

PONE-D-20-25142R2

Establishing the Content of Gender Stereotypes Across Development

Dear Dr. Sullivan:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Alexander N. Sokolov

Academic Editor

PLOS ONE