Survey data of social, emotional, and behavioral skills among seven independent samples

Abstract

The data presented in this article— originally reported by Soto and colleagues (Soto et al., in press)— assess social, emotional, and behavioral (SEB) skills, indexed by the Behavioral, Emotional, and Social Skills Inventory (BESSI), across seven independent samples (N = 6,309). Four of the datasets (N = 5000) were collected using an online survey housed on PersonalityLab.org. In two of these internet datasets, participants provided their responses to sociodemographic items, subsets of BESSI items (45 – 102 items), and the Big Five Inventory-2 (BFI-2, 60 items). In the other two internet datasets, participants answered the same sociodemographic items and the full BESSI questionnaire (192 - 200 items). The fifth dataset was collected via an online survey sponsored by the Character Lab Research Network and included responses from 499 high school students. The High School Student Sample completed sociodemographic items, the full BESSI (192 items), and measures of academic engagement, occupational interests, peer acceptance, friendship quality, romantic relationship satisfaction, family relationship satisfaction, volunteerism, physical exercise, and life satisfaction (96 total items). The sixth dataset was collected using the Qualtrics Online Sample service, and 488 adult respondents completed an extended, observer-report version of the BESSI (284 items), sociodemographic items, and information regarding their relationship with the person whom they were reporting on (7 items). The seventh data set consisted of college students (N = 322) from Colby College. The College Student Sample completed a survey on Qualtrics that included sociodemographic items, the full BESSI (192 items), the BFI-2 (60 items), and four other SEB skill inventories (116 items). All datasets, questionnaires, and scoring forms are hosted on OSF. The data can be used to (1) understand the structure and organization of SEB skills, (2) model the relationship between SEB skills and conceptually adjacent constructs such as personality traits and character strengths, and (3) explore the associations between SEB skills and consequential outcomes.

Specifications Table

Subject	Psychology
Specific subject area	Personality Psychology; Educational Psychology; Measurement and Assessment; Socioemotional Learning; Noncognitive Skills
Type of data	Raw data in excel files and .sav files and analyzed data in Tables.
How the data were acquired	All data were collected via online surveys and converted into the .xlsx format and .sav for formal analysis in SPSS or R. Copies of each survey and scoring information are provided in the supplemental materials.
Data format	Raw Analyzed
Description of data collection	The surveys were all administered to participants between the ages of 10 and 80 online on either PersonalityLab.org (N = 5000) or Qualtrics (N = 1309). Various exclusion criteria, described in detail below, were used to ensure data quality.
Data source location	Internet Sample A - D Datasets: North America, Australia, Europe; Observer Sample Dataset: North America; High School Student Sample Dataset: North America, United States; College Student Sample Dataset: Colby College, Waterville, Maine, U.S., 44.5639° N, 69.6626° W
Data accessibility	All datasets, surveys, and scoring forms can be found on OSF. Repository Name: open science framework; Direct URL to Internet Sample A dataset: https://osf.io/n4r8a/ for xlsx and https://osf.io/kczvd/ for sav; Direct URL to Internet Sample B dataset: https://osf.io/c5b3p/ for xlsx and https://osf.io/7u2qe/ for sav; Direct URL to Internet Sample C dataset: https://osf.io/4rftv/ for xlsx and https://osf.io/qgzus/ for sav; Direct URL to Internet Sample D dataset: https://osf.io/ewpxu/ for xlsx and https://osf.io/utq27/ for sav; Direct URL to Observer Sample dataset: https://osf.io/uqxd7/ for xlsx and https://osf.io/6qhez/ for sav; Direct URL to College Student Sample dataset: https://osf.io/7cktv/ for xlsx and https://osf.io/p7qk5/ for sav;
	Direct URL to High School Student Sample dataset: https://osf.io/9xqws/ for xlsx and https://osf.io/h7txz/ for sav
Related research article	C.J. Soto, C.M. Napolitano, M.N. Sewell, H.J. Yoon, B.W. Roberts, An integrative framework for conceptualizing and assessing social, emotional, and behavioral skills: The BESSI. J. Personal. Soc. Psychol., In Press.

Value of the Data

• •The data can be used to explore the structure of social, emotional, and behavioral (SEB) skills across multiple developmental periods (adulthood and adolescence) and assessment perspectives (self-report and observer-report).
• •The data presented here can be used to investigate associations among SEB skills and other related constructs such as personality traits, socioemotional competencies, character, and developmental strengths in an adolescent sample.
• •Practitioners in education and youth development contexts can utilize this data to develop a measurement protocol to assess SEB skills in youth.
• •Psychology and education researchers can benefit from this data by exploring how SEB skills promote positive developmental outcomes in youth and using these insights to further research the development, antecedents, and consequences of SEB skills across contexts and across the lifespan.

1. Data Description

The aims of this data collection were to (1) build an integrative model of social, emotional, and behavioral (SEB) skills, (2) investigate the nomological network of SEB skills by examining their relationship with related constructs, and (3) explore how SEB skills relate to academic achievement and engagement, well-being, occupational interests, and social relationships among adolescents [1]. The seven independent samples (N = 6,309) presented in this article were collected between January of 2019 and May of 2021. We describe each dataset in detail below in the order in which they were collected.

1.1. Internet samples

All participants in the Internet Sample A, B, C, and D datasets completed a survey titled “Find Your Hogwarts House” on PersonalityLab.org, a non-commercial website. All participants were asked to provide responses to four demographic items assessing their gender, age, race and ethnicity. Participants in Internet Sample A and B completed the Big Five Inventory-2 (BFI-2) [2]. The BFI-2 consists of 60 items that measure the Big Five personality domains—Extraversion, Agreeableness, Conscientiousness, Negative Emotionality, and Open-Mindedness [2]. Participants in Internet Sample A and B indicated whether the BFI-2 statements were descriptive of them on a 5-point agreement scale (1 = disagree strongly, 5 = agree strongly). Finally, all internet samples completed a version of the Behavioral, Emotional, and Social Skills Inventory (BESSI), a measure of SEB skill facets and domains [1]. The data from each Internet Sample is described in more detail in the following paragraphs.

Internet Sample A. The Internet Sample A dataset (N = 2000) was collected in January and February of 2019. Participants in Internet Sample A completed one of five subsets of BESSI items. Each subset focused on a specific SEB skill domain—Social Engagement (58 items), Cooperation (59 items), Self-Management (79 items), Emotional Resilience (45 items), or Innovation Skills (71 items)—and instructed participants to rate how easy or hard it was for them to perform each behavior well (1 = very hard, 2 = pretty hard, 3 = neutral; in between, 4 = pretty easy, to 5 = very easy). The surveys used for Internet Sample A can be found on OSF [Social Engagement (https://osf.io/dk4pq/), Cooperation (https://osf.io/bcwps/), Self-Management (https://osf.io/vfnh3/), Emotional Resilience (https://osf.io/7yutz/), and Innovation (https://osf.io/cva5x/)].

Internet Sample B. The Internet Sample B dataset (N = 2000) was collected from March to May of 2019. Participants in Internet Sample B also completed one of five subsets of BESSI items—Social Engagement (102 items), Cooperation (100 items), Self-Management (99 items), Emotional Resilience (98 items), or Innovation Skills (99 items)—and these subsets used the same rating scale described above. The surveys used for Internet Sample B can be found on OSF [Social Engagement (https://osf.io/7j9y3/), Cooperation (https://osf.io/95v8q/), Self-Management (https://osf.io/b7mw4/), Emotional Resilience (https://osf.io/8vca6/), and Innovation (https://osf.io/4hgb9/)].

Internet Sample C. The Internet Sample C dataset (N = 400) was collected from August to November of 2019. Participants in Internet Sample C completed a complete version of the BESSI (200 items). This version of the BESSI asked participants to rate how well they could perform each behavior, reflecting their current level of expertise [1 = not at all well (beginner), 2 = not very well (advanced beginner), 3 = pretty well (intermediate), 4 = very well (advanced), 5 = extremely well (expert)]. The rating scale was adapted to a “how-well” scale to focus on skill level rather than effort. The survey used for Internet Sample C can be found on OSF [https://osf.io/3dw78/].

Internet Sample D. The Internet Sample D dataset (N = 600) was collected between November of 2019 and March of 2020. Participants in Internet Sample D completed a full version of the BESSI (192 items). This version asked participants to rate how well they could perform each behavior without the expertise labels (e.g., 1 = not at all well, 5 = extremely well). The survey used for Internet Sample D can be found at OSF [https://osf.io/pndh9/]. Table 1 provides the demographic characteristics of Internet Samples A–D.

Table 1.

Demographics of Internet Samples A–D.

Variables	Internet Sample A	Internet Sample B	Internet Sample C	Internet Sample D
Age Mean (Std. Deviation)	24.17(9.68)	23.31(9.13)	25.23(9.90)	26.41(9.10)
Gender
Male	49.0%	50.0%	50.0%	50.0%
Female	49.0%	50.0%	50.0%	50.0%
Another gender	2.0%	0.0%	0.0%	0.0%
Ethnicity
White/Caucasian	71.2%	71.1%	67.8%	73.5%
Black/African American	2.9%	3.9%	5.3%	2.3%
Hispanic/Latino	10.3%	9.4%	8.3%	8.8%
Asian/Asian American	13.4%	12.9%	13.8%	13.5%
American Indian/Native American	2.1%	2.8%	2.5%	1.0%
Native Hawaiian/Pacific Islander	0.8%	1.2%	0.8%	0.5%
Another ethnicity	5.9%	5.3%	7.5%	4.8%
Multiple ethnicities	9.8%	8.6%	8.0%	7.0%
Declined to report	4.5%	4.3%	4.5%	3.7%
Nationality
United States	48.5%	54.0%	57.0%	55.3%
United Kingdom or Ireland	4.5%	2.9%	3.0%	4.2%
Canada	4.0%	3.7%	3.5%	3.7%
Australia or New Zealand	3.2%	4.1%	2.5%	1.7%
Another country	22.7%	19.3%	18.0%	18.3%
Declined to report	17.1%	16.0%	16.0%	16.8%
Total sample size (N)	2000	2000	400	600
BESSI items completed	45-79	98-102	200	192

Note. All values are frequencies unless otherwise noted in the variables column.

1.2. High school student sample

The High School Student Sample (N = 499) contains data from students from large suburban and urban high schools across the United States. Data was collected between December of 2019 and February of 2020 on the Qualtrics platform. Participants completed five demographic items assessing their gender, age, race and ethnicity, and grade in school. Participants completed the same version of the BESSI as Internet Sample D and the College Student Sample (192 items, how-well rating scale). Participants also completed a battery of 96 items assessing academic, relationship, and well-being outcomes.

Academic engagement was assessed using a 12-item version of the Engagement versus Disaffection with Learning measure (EDL) [3]. Occupational interests were assessed using the O*Net Mini Interest Profiler (Mini-IP) [4]. School-reported grades, aggregated into an overall grade point average, were available for a subsample of participants (N = 474) for winter and spring courses.

Peer acceptance was assessed using two items: One item asked participants to rate their level of popularity among their peers on a 9-point scale [5], and the other item asked participants to rate their level of status within their social group on a 9-point scale [6]. Friendship quality with the participants’ best friend was assessed using a 15-item version of the Friendship Qualities Scale (FQS) [7]. Participants currently involved with a romantic partner for at least one month completed two items adapted from the Dyadic Adjustment Scale (DAS) to assess the degree of happiness in the relationship and feelings about the future of the relationship [8]. The quality of participants’ relationships with their parents was assessed using six items adapted from the Dunedin Study of lifespan development [9].

Participants responded to four items assessing their current, past, and prospective volunteering activities [10]. Participants answered three items from the Godin Leisure-Time Exercise Questionnaire (GLTEQ) indicating the number of days in a typical week that they engage in strenuous, moderate, and mild physical exercise [11]. Finally, participants completed the 5-item Satisfaction With Life Scale [12]. Table 2 provides the demographic characteristics for the High School Student Sample. Survey materials can be found at OSF [https://osf.io/vdrwa/].

Table 2.

Demographics of high school sample.

Variables	% / Mean (Std. Deviation)
Age	15.62(1.17)
Gender
Male	45.1%
Female	52.7%
Another gender	1.2%
Not reported	1.0%
Ethnicity
White/Caucasian	42.1%
Black/African American	34.1%
Hispanic/Latino	18.4%
Asian/Asian American	3.0%
Another ethnicity	0.6%
Multiple ethnicities	4.4%
Not reported	4.2%
Grade Level
9th grade	30.9%
10th grade	33.5%
11th grade	20.0%
12th grade	14.6%
Not reported	1.0%
Total sample size (N)	499
BESSI items completed	192

1.3. Observer sample

The Observer Sample consists of data from adult participants (N = 488), recruited through the Qualtrics Online Sample service, collected in April of 2020. Participants completed four demographic items assessing their gender, age, and race and ethnicity. Participants were asked to provide detailed and accurate ratings of another person's behavior. Across seven items, participants provided demographic information about their chosen target, details on their relationship with the target (e.g., friends, spouse, family, etc.), how well they know the target (1 = not at all, 5 = extremely well), how much they like the target (1 = strongly dislike, 5 = strongly like), and how long they've known the target. Finally, participants completed an observer-report version of the BESSI (284 items) using the how-well rating scale with expertise labels [1 = not at all well (beginner), 5 = extremely well (expert)]. Table 3 provides the demographic characteristics of the Observer Sample. Survey materials can be found at OSF [https://osf.io/5zpxt/].

Table 3.

Demographics of observer sample and target characteristics.

Variables	% / Mean (Std. Deviation)
Age	49.25(16.32)
Gender
Male	51.6%
Female	48.0%
Another gender	0.4%
Ethnicity
White/Caucasian	84.2%
Black/African American	6.1%
Hispanic/Latino	3.1%
Asian/Asian American	4.7%
Another ethnicity	1.8%
Target Relationship
Romantic partner	41.2%
Friend	27.0%
Family member	22.5%
Coworker	4.9%
Other acquaintance	4.3%
Length of relationship	24.37 (17.86)
Knowledge of Target
Not at all	.4%
A little bit	2.3%
Pretty well	9.4%
Very well	22.5%
Extremely well	65.4%
Feelings toward Target
Strongly dislike	2.0%
Dislike a little	3.7%
Neutral	4.1%
Like a little	10.2%
Strongly like	79.9%
Total sample size (N)	488
BESSI items completed	284

Note. Length of relationship is in years.

1.4. College student sample

The College Student Sample dataset (N = 322) contains data from students at Colby College, a private college in the Northeastern United States and was collected between November of 2019 and May of 2021 on the Qualtrics platform. Participants completed five demographic items assessing their gender, age, race and ethnicity, and classification (e.g., first-year, sophomore, etc.). Participants completed the same version of the BESSI as Internet Sample D (192 items, how-well rating scale). Participants also completed the BFI-2, and the Positive Youth Development Short-Form (PYD-SF) [13]. The PYD-SF is a 34-item measure of developmental strengths across five domains: Competence, Confidence, Character, Caring, and Connection. Twenty-six of its items are pairs of contrasting statements that participants rated on a 5-point semantic differential scale (e.g., 1 = left is really true, 5 = right is really true). The remaining eight items are individual statements or activities that participants rated on a 5-point scale reflecting their degree of agreement, importance, or descriptiveness.

A subsample of participants (N = 249) also completed a measure of character strengths—the Tripartite Taxonomy of Character (TTC) [14] —and two measures of socioemotional competencies—the Social and Emotional Competency Assessment (SECA) [15], and Social-Emotional Learning Surveys (SELS) [16]. The TTC assesses Interpersonal, Intrapersonal, and Intellectual strengths across 24 items on a 7-point frequency scale (e.g., 1 = almost never, 7 = almost always). The SECA uses 40 items to assess Self-Management, Social Awareness, Relationship Skills, and Responsible Decision-Making on a 4-point scale indicating how difficult or easy it is for the respondent to enact that skill. The SELS consists of 18-items measuring Self-Management, Social Awareness, Self-Efficacy, and Growth Mindset. All items are on a five-point scale gauging either frequency, degree of confidence, agreement, or an item-specific outcome. Table 4 provides the demographic characteristics for the College Student Sample. Survey materials can be found at OSF [https://osf.io/8rxbk/].

Table 4.

Demographics of college sample.

Variables	% / Mean (Std. Deviation)
Age	19.38(1.13)
Gender
Male	29.8%
Female	69.6%
Another gender	0.6%
Ethnicity
White/Caucasian	74.5%
Black/African American	6.5%
Hispanic/Latino	8.7%
Asian/Asian American	20.2%
Another ethnicity	2.2%
Multiple ethnicities	11.8%
Classification
First-years	40.4%
Sophomores	27.6%
Juniors	21.4%
Seniors	10.6%
Not reported	1.0%
Total sample size (N)	322
BESSI items completed	192

2. Experimental Design, Materials and Methods

Because the datasets presented here were recruited via multiple means, we present each dataset or set of datasets in turn. Internet Samples A–D were recruited passively, and participants could find the “Find your Hogwarts House” survey through search engines, links on other websites, and word of mouth. After completing the survey, participants were provided free feedback on their compatibility with the four Hogwarts Houses (i.e., Gryffindor, Hufflepuff, Ravenclaw, and Slytherin) featured in the Harry Potter series [17]. This framing was designed to motivate participants to respond accurately so they could receive meaningful feedback.

The Observer Sample participants were recruited using the Qualtrics Online Sample service and were paid approximately $3 USD for their participation. The College Student Sample participants were actively recruited from psychology classes at Colby College. The College Student Sample participants received course credit in exchange for their participation. Finally, the High School Student sample participants were recruited from large urban and suburban high schools that were a part of the Character Lab Research Network, a consortium of schools across the country working collaboratively with scientists to advance scientific insights that help kids thrive [18].

2.1. Exclusion criteria and missing data

To ensure data quality, some participants across the samples were excluded from analyses. In the Internet Samples A–D datasets, participants were excluded if they (1) answered fewer than 90% of the survey questions, (2) did not report their gender, (3) did not report their age, or reported an age of less than 10 or greater than 80 years old, (4) reported that they did not speak English fluently, (5) reported inconsistent information about their nationality, or (6) completed the survey multiple times (as indicated by duplicate submissions from the same IP address). For Internet Samples A and B, participants who had a within-person standard deviation of less than 0.50 across the BFI-2 items were also excluded from analyses. After applying these exclusion criteria, the final sample was randomly selected to include an equal number of male and female participants.

In the Observer Sample, College Student Sample, and High School Student Sample, participants were excluded from analysis if (1) they answered fewer than 90% of the BESSI items, (2) failed either of two attention-check items, and (3) completed the survey in less than one-third of the median time. In addition, participants in the Observer Sample were excluded if they failed to answer either of two open-ended comprehension items, reported inconsistent demographic information, or reported that they did not answer the survey items honestly. Participants in the College Student Sample were also excluded if they submitted the survey multiple times.

Our exclusion criteria removed a total of 3,881 raw cases from Internet Sample A, 4,222 cases from Internet Sample B, 2,635 cases from Internet Sample C, 4,104 cases from Internet Sample D, 52 cases from the Observer Sample, 42 cases from the College Student Sample, and 674 cases from the High School Student Sample. After these cases were removed from the final samples, any missing item responses were replaced with the item mean, rounded to the nearest possible response.

2.2. Data analyzed

SEB Skill Facets and Domains. In Internet Samples C and D, as well as the College and High School Student Samples, composite scores for the 32 BESSI skill facets were computed as the mean response to the six items included on each facet scale. Similarly, composite scores for the five higher-order skill domains were computed as the mean of the facet scales assigned to each domain, with half weight given to interstitial facets that were assigned to two skill domains. SPSS and R syntax for computing BESSI domain and facet scores are available on OSF [https://osf.io/c8uq7/ and https://osf.io/mt8yv/, respectively]. Table 5 demonstrates the organization of the BESSI skill facets and domains.

Table 5.

BESSI skill domain and facet organization.

Social Engagement	Cooperation	Emotional Resilience	Self-Management	Innovation	Interstitial	Compound
Leadership skill	Teamwork skill	Stress regulation	Task management	Abstract thinking skill	Energy regulation	Adaptability
Persuasive skill	Capacity for trust	Capacity for optimism	Time management	Creative skill	Ethnical competence	Capacity for independence
Conversational skill	Perspective-taking skill	Anger management	Detail management	Artistic skill	Impulse regulation	Self-reflection skill
Expressive Skill	Capacity for social warmth	Confidence regulation	Organizational skill	Cultural competence	Information processing skill
			Responsibility management
			Capacity for consistency
			Goal regulation
			Rule following skill
			Decision-making skill

Note. Interstitial skills equally correspond to two skill domains: Energy regulation = Self-Management + Social Engagement; Ethical competence = Self-Management + Cooperation; Impulse regulation = Self-Management + Emotional Resilience; Information process skill = Self-Management + Innovation. Compound skills correspond to three or more domains.

Other Social, Emotional, and Behavioral Constructs. The BFI-2 personality traits, PYD developmental strengths, SECA and SELS socioemotional competencies, and TTC character strength measures were scored by computing all scale and subscale scores as a simple mean item response (after reverse-coding any negatively keyed items).

High School Student Sample Outcomes. Academic engagement was computed as the mean response to the 12 EDL items (after reverse-coding the disaffection items). Occupational interests domains were computed as the mean response to the five items corresponding to each domain measured by the Mini-IP. Peer acceptance and life satisfaction were scored by computing the mean of their scale items.

Indexes of parental relationship satisfaction with the participants’ mother and father were created from the average of three standardized items. An overall score of friendship quality was computed by averaging the 12 FQS items from the companionship, help, security, and closeness subscales. Romantic relationship satisfaction was computed from the average of two standardized DAS items.

Volunteering behavior items were submitted to a principal components analysis, as described by Carlo and colleagues [10], and scores on the first unrotated component were saved as an overall index of volunteer behavior. An overall indicator of exercise behavior was computed as a weighted composite of the three GLTEQ items, as described by Godin and Shephard [11].

Ethics Statements

Informed consent or assent was obtained from all participants in the College, Observer, and High School Samples before data collection. In the High School Sample, parents could complete an opt-out form to exclude their children from participating. A waiver of informed consent was granted for Internet Samples A–D. Internet Samples A–D and the College Student Sample data collection and procedures were approved by the Colby College Institutional Review Board (Protocol #2019-001 & #2019-083, respectively). The Observer Sample data collection and procedures were approved by the University of Illinois at Urbana-Champaign Office for the Protection of Research Subjects (Protocol #20097). The High School Sample data collection and procedures were approved by the Advarra Institutional Review Board (Protocol #MOD00495112) and were conducted by the Character Lab Research Network.

CRediT authorship contribution statement

Madison N. Sewell: Formal analysis, Writing – original draft. Christopher J. Soto: Conceptualization, Methodology, Formal analysis, Data curation, Writing – review & editing. Christopher M. Napolitano: Conceptualization, Methodology, Formal analysis, Writing – review & editing. Hee J. Yoon: Data curation, Writing – review & editing. Brent W. Roberts: Conceptualization, Methodology, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.