ArticlePDF Available

Use of LinkedIn Data and Machine Learning to Analyze Gender Differences in Construction Career Paths

August 2022
Journal of Management in Engineering 38(6)

August 2022
38(6)

DOI:10.1061/(ASCE)ME.1943-5479.0001087

Authors:

Paul J. Hickey

University of Maryland, College Park

Abdolmajid Erfani

Michigan Technological University

Qingbin Cui

University of Maryland

Will women and men follow distinctively different paths to achieve executive engineering leadership positions in the US architecture, engineering, and construction (AEC) industry? Using Engineering News Record’s (ENR’s) 2019 Top 400 list, this research analyzed LinkedIn profiles for over 2,800 executives to assess career differences between genders. Statistical comparisons of important features, highlighted by number of companies, titles, education, and network size, revealed a significant impact of gender on individual career paths. A key finding was that men ascend to leadership with a single firm throughout their career, outpacing women almost fourfold (37% to 10%). Applying random forest (RF) as an ensemble classifier, researchers successfully predicted profile gender with accuracy of 98.95% for training and 89.53% for testing samples. Collating and categorizing the activities and milestones of individual and collective executives offer insight regarding successful experiences, skills, and choices to reach leadership roles. This creates a roadmap for current and future early and midlevel professionals to model their own vocational journey and accelerate progression up the corporate ladder. From an industry perspective, firms deprive themselves and customers of the proven wide-ranging benefits of diversity.

Data set career path feature visualization: (a) number of career companies; (b) years to current job; (c) network size; (d) company size; and (e) highest level of education earned.

…

Feature importance sorting.

…

Number of career companies. Fig. 4. Network size.

…

Number of career positions.

…

Average number of positions held by company revenue.

…

Figures - uploaded by Paul J. Hickey

Content may be subject to copyright.

Content uploaded by Paul J. Hickey

Content may be subject to copyright.

Use of LinkedIn Data and Machine Learning to Analyze

Gender Differences in Construction Career Paths

Paul J. Hickey1; Abdolmajid Erfani2; and Qingbin Cui, A.M.ASCE3

Abstract: Will women and men follow distinctively different paths to achieve executive engineering leadership positions in the US

architecture, engineering, and construction (AEC) industry? Using Engineering News Record’s (ENR’s) 2019 Top 400 list, this research

analyzed LinkedIn profiles for over 2,800 executives to assess career differences between genders. Statistical comparisons of important

features, highlighted by number of companies, titles, education, and network size, revealed a significant impact of gender on individual

career paths. A key finding was that men ascend to leadership with a single firm throughout their career, outpacing women almost fourfold

(37% to 10%). Applying random forest (RF) as an ensemble classifier, researchers successfully predicted profile gender with accuracy of

98.95% for training and 89.53% for testing samples. Collating and categorizing the activities and milestones of individual and collective

executives offer insight regarding successful experiences, skills, and choices to reach leadership roles. This creates a roadmap for current and

future early and midlevel professionals to model their own vocational journey and accelerate progression up the corporate ladder. From an

industry perspective, firms deprive themselves and customers of the proven wide-ranging benefits of diversity. DOI: 10.1061/(ASCE)

Author keywords: Construction industry; Leadership; Gender; Diversity; LinkedIn; Social media networking.

Introduction

Historical perspective on diversity traditionally focused on legal

and moral perspectives (Grant and Kleiner 1997). Coupled with the

Kennedy administration’s legislative efforts in the early 1960s, the

civil rights movement propelled women’s rights into the forefront

of social consciousness, driving change in educational and profes-

sional outcomes (Jardina and Burns 2016). This paradigm change

allowed women to pursue opportunities previously unavailable,

creating a significant shift in the late 20th-century workforce. The

passing of the Civil Rights Act of 1964 was a key milestone in US

society, seeking to end discrimination based on an individual’s

race, color, religion, sex, or national origin (US Equal Employment

Opportunity Commission 2021). Nonetheless, even laws and regu-

lations accompanying this landmark legislation failed to generate fair

balance. As a result, affirmative action programs emerged to bridge

the gap and eliminate the imbalances. With the continuing growth

of women in the workforce and associated greater prominence of

gender diversity, policies must continue to evolve to address the

21st-century workplace (Byrd and Scott 2014). Despite greater pub-

lic visibility and scrutiny, women’s architecture, engineering, and

construction (AEC) representation still lags behind general US pop-

ulation distribution. Even in the engineering field, women gravitate

towards disciplines offering societal impact (Dzombak and Mehta

2017). Targeted efforts to increase participation by women only gen-

erated modest gender diversity improvements in select engineering

subsegments (National Academy of Engineering 2018). Specifically

in the civil and construction engineering field, programs yield only

minimal market penetration where women fill approximately 1 in

10 industry positions and fewer than 1 in 25 executive engineering

leadership billets (Hickey and Cui 2020).

This investigation further quantified hierarchal organizational

imbalance in AEC companies, adding a detailed analysis of career

path components. Although significant research exists on gender

studies, with an accelerating corpus focused on the civil engineer-

ing and construction industry, authors found a lack of data-driven

measurement of career paths, underlying professional decisions,

and comparison between women and men. Today’s early and mid-

level career women lack role models to inspire them for continued

growth, hindering their ability to achieve future leadership posi-

tions (Roebuck and Smith 2013). Further exacerbating the issue,

a recent study found that women supervisors leave their jobs more

often than men (Maurer et al. 2021). In an effort to identify factors

to offset this imbalance, authors extracted profiles of leading indus-

try firms’executive teams to provide roadmaps of successful ac-

tions and behaviors. With over 774 million users from more than

200 countries and territories, LinkedIn consolidates the largest cor-

pus of professional experience. When considering the accuracy of

self-populated career information, authors considered the adverse

professional ramifications resulting from discovered falsehoods

embedded in profiles. Therefore, authors deemed the published

LinkedIn information peer reviewed and accurate. However, reli-

ance on full disclosure or propensity to inflate their accomplish-

ments by all candidates offers study limitations and creates

potential for risk in the data set. As part of the data integrity process,

applied statistical testing found no difference in reporting detail or

completeness between genders. Across the larger workspace, focus

on professional social media ethical practices relating to accuracy,

validity, and documentation continues to grow (Kettunen and

Makela 2019). As detailed in the methodology section, a further

step to maintain accuracy included data cleaning analysis to exclude

1Ph.D. Student, Dept. of Civil and Environmental Engineering, Univ. of

Maryland, College Park, MD 20742. Email: phickey1@umd.edu

2Ph.D. Student, Dept. of Civil and Environmental Engineering, Univ. of

Maryland, College Park, MD 20742. ORCID: https://orcid.org/0000-0002

-4703-1248. Email: erfani@umd.edu

3Professor, Dept. of Civil and Environmental Engineering, Univ. of

Maryland, College Park, MD 20742 (corresponding author). Email: cui@

umd.edu

Note. This manuscript was submitted on October 24, 2021; approved on

May 19, 2022; published online on August 3, 2022. Discussion period open

until January 3, 2023; separate discussions must be submitted for individual

papers. This paper is part of the Journal of Management in Engineering,

J. Manage. Eng., 2022, 38(6): 04022060

unmaintained profiles containing minimal data. A strategic objec-

tive of this study investigated the potential impact of gender on indi-

vidual career paths using machine learning (ML) techniques and

statistical tests considering important career path features extracted

from LinkedIn data of construction leaders’profiles.

The authors recognize the ongoing evolution regarding gender

identity and its importance in society. For the purposes of this

study, we applied a binary definition of female and male in compil-

ing, analyzing, and reporting gender.

Literature Review and Hypothesis Development

Prior to extracting profile data, the authors drafted an initial set of

research questions and hypotheses considering prior academic the-

ories. After data collection and cleaning, researchers applied social

capital theory and social cognitive career theory (SCCT) to the

resultant LinkedIn data set. Examination includes primary features

of individual career paths such as number of firms and positions,

highest level of education earned, and professional network size.

Koch et al. (2017) found that career-minded individuals who

maintain a singular focus on work achieve executive levels faster

and at higher rates than counterparts who seek a balance between

work and personal lives. Because women traditionally shoulder a

greater responsibility in family responsibilities, they experience a

significant disadvantage due to competing obligations. In fact, US

women head more single-parent households than men and experi-

ence a higher risk of dropping out of the workforce after childbirth

(Aisenbrey and Fasang 2017). As a result of these external influ-

ences, professional women temporarily exited the workforce at

higher rates, 37% (43% for women with children) to 24% for men

(Hewlett and Luce 2005). Although most of these women, 93%,

planned on resuming their vocation, gaps disrupted their career arc.

Even high-profile women left key positions, including CEOs of

Fortune 500 companies, to focus on family, both aging parents

and young children (Hewlett and Luce 2005). Ultimately, these

contributing factors create career drag, inhibiting long-term pro-

gression, especially compared to their male colleagues.

Throughout a person’s career, they continually redefine and

reassess the relationship between work and personal lives. Prior

research documents women’s greater desire for work–life balance

(Koch et al. 2017). That equilibrium underlies women citing family

and unsatisfying jobs as key reasons for leaving the workforce,

whereas men’s predominant motivations included switching ca-

reers, obtaining greater levels of education, or starting a business

(Hewlett and Luce 2005). When viewed from the company per-

spective, gender-diverse leadership teams create cultures offering

policies that create better work–life balance (Baker et al. 2021).

However, despite recent progress and public commitment to

change, most AEC firms fail to achieve fair implementation in prac-

tice (Hasan et al. 2021).

Social Capital Theory

Social capital represents the theory of capturing resources residing

in social relationships and leveraging them into better business out-

comes (Lin 2001). Social networks generate benefits, including

physical and monetary resources, social support, and access to in-

formation (Alaa et al. 2018). Since its inception, researchers have

dedicated significant research to social network theory and devel-

oped divergent concepts to the topic (Brass 2003). In their 2001

study, Seibert, Kraimer, and Liden tested social capital theory, ap-

plying it against networks and relationships, assessing the resultant

impact on career progression. Two theories emerged: weak ties—

connections outside the core group and structural holes—where

direct connections of the individual remain separate from each

other (Seibert et al. 2001). Follow-on studies identified strong ties,

which generate bonding social capital, whereas bridging social

capital emerges from weak ones (Alaa et al. 2018;Neumeyer et al.

2019). These otherwise disparate network combinations from

bridging social capital offered unique information unavailable to

the other members of the direct group, often leading to opportuni-

ties for new jobs. Overall, Seibert’s(Seibert et al. 2001) study found

that both theories positively relate to a person’s accessibility to so-

cial resources and resultant career progression.

All interactions between people require communication, whether

verbal, written, or other methods. Resultant social networks depend

more on the structural links rather than the nodes, for example, re-

lationships vice the individual players (Brass 2003). Over time, pro-

longed interaction between individuals establishes the structure of

relationships and resultant bonding social capital within an organi-

zation (G´orska et al. 2022). In a hierarchal configuration, the basis

for some relationships is anchored in positional hierarchy, whereas

others grow organically. Ultimately, these collective personal behav-

iors, coupled with the overarching structure, impact the overall

network. Add that each new generation of incoming talent chal-

lenges the status quo and entrenched leadership’s ability to maintain

existing traditions. Recent focus on social justice, including gender

equality, promises to create a lasting impact on future direction of

organizations.

Even in the early 1980s, experts recognized the need to access

leadership talent from previously untapped groups, including

women and minorities (Kanter 1981). Conversely, selection of can-

didates like oneself reflects homosocial reproduction and perpetu-

ates the status quo (Kanter 1993). This homophily, preference of

one’s own type, strengthens the bonding social capital within the

core group (Alaa et al. 2018). With men filling most senior positions

(Hickey and Cui 2020), this provides opportunities to exercise

self-similar preference and hire or promote the next generation of

industry leaders (Elliott and Smith 2004). Reversing this trend and

generating improvements in diversity requires organizations to re-

flect upon and change their hiring and promotion practices. Specifi-

cally, building formal and informal workplace networks offers

opportunities for women to strengthen bridging social capital across

functional areas and externally from the firm (G´orska et al. 2022).

As society continues to evolve its perceptions on equity, down-

stream impact to industry follows. Kanter (2021) coined the term

“advanced”leadership to describe evolving change beyond the tra-

ditional hierarchal management to encompass broad cross func-

tional and multisector operations and interactions. Formalization

of these concepts further emphasizes the need for firms to recognize

this seismic change for future success. An opportunity for compa-

nies to positively contribute to forwarding diversity, transitioning

from vague goals and objectives to a specific, clearly defined vision,

increases emphasis on the importance and group focus (Kanter

1981). Expanded understanding of bonding and bridging social

capital offers insight for firms to provide equal opportunities for all

employees (G´orska et al. 2022). This supports findings from this

paper’s authors’prior study on executive suite leadership (Hickey

and Cui 2020). Results found that a public commitment to diversity,

defined as prominent publication of inclusion on a company’s

website, resulted in higher women’s representation.

Although this LinkedIn investigation concentrates on the senior

and executive organizational levels, some prior studies focused

on impact of corporate board diversity. Arguably, each group offers

a different impact on the daily operation of the company, where

boards may only meet a few times per year. However, Kanter’s

(1993) homosocial reproduction theory positions a diverse board

to disrupt perpetuating hiring and promoting similar individuals,

J. Manage. Eng., 2022, 38(6): 04022060

specifically white men. Despite proven benefits of diversity

(Sunindijo and Kamardeen 2017), a study of board diversity and

firm reputation found that the impact of women on boards created

positive impacts on service industries and negative ones for pro-

ducing sectors, including construction (Brammer et al. 2009).

Brammer’s results aligned with the proximity linkage between the

company and end users. Therefore, AEC’s insulation from the gen-

eral public contributes to its slower pace of inclusion compared

to other industries. Another study found gender diversity on boards

had a positive relationship to greater organizational innovation

(Miller and Del Carmen Triana 2009). However, a consideration

of possible lower impact of women on boards resides in the fact that

women hold fewer chairperson or committee lead positions (Miller

and Del Carmen Triana 2009). Other research identified corporate

social responsibility (CSR), a measurement of self-regulating busi-

ness model for social accountability, increases with inclusion of

women on boards (Zhang et al. 2013).

Number of Career Firms

Researchers considered the impact of internal and external influen-

ces on women’s and men’s respective career arcs. Even with the

significant increase in female entry into the workforce, turnover

for women exceeds that of men (Yousaf et al. 2014). Across all

industries, women experience higher instances of separation, in-

cluding both voluntary and involuntary (Frederiksen 2008). This

study assesses whether this trend would apply to the AEC industry

and women reaching equivalent organizational levels with fewer

career changes. Evaluating career mobility provides insight whether

individuals achieve greater success, as defined in this study by

reaching leadership positions, either by remaining with a single

organization or switching to multiple companies. Inherently un-

equal, organizations exhibit increasingly strong social capital with

higher levels of leadership (Alaa et al. 2018). Limited accessibility

to these assets creates disadvantages for women in their progression

to the executive suite. Tight organizational bonding social capital

promotes the status quo and hinders minority groups from progress-

ing, so they must engage in bridging activities to reach their full

potential in leadership (G´orska et al. 2022).

Short term promotions and pay increases often drive decisions

to join a new firm, but will long-term career progression benefit

as well? Contrary to perceptions that company changes lead to

accelerated growth, most companies continue to select internal

candidates who already possess significant networks and a deep

understanding of the company philosophy (Koch et al. 2017). This

concentrated focus on bonding social capital, whether conscious or

not, favors homophily and selection of those mirroring the current

leadership team. As an extension, Koch et al. (2017) and Hamori

and Kakarika (2009) found that CEOs with more frequent job

changes and less time served at their current company took longer

to reach the top position, with the insider track presenting the more

frequent pattern to reach executive leadership, remaining with one

company for much of their career.

H1: Male-dominated executive teams’bonding social capital

perpetuate the status quo, limiting opportunities for women, caus-

ing them to change firms more often during their career progression

to leadership.

Professional Network Size

The authors considered the impact of accumulating industry con-

nections, both internally and externally to their companies, on

women’s and men’s respective career arcs. Women generate greater

links and associated contacts (Roebuck and Smith 2013). This

study assessed whether this benefits their growth within the indus-

try. Professionally, networking embodies the specific pursuit of ac-

cumulating relationships to benefit one’s career. Although it is

traditionally associated with job seeking, effective implementation

requires continuous application of this practice throughout a career

(Eason and Krause 2001). Creating a diverse group of professional

advocates offers a support group to seek advice, gain insight, and

access opportunities. Proven to expedite career progression and in-

crease earnings, job performance, and career satisfaction, network-

ing offers an important tool for anyone in the workforce (Davis

et al. 2020). The explosion of social networking shows the potential

benefits of electronic media, highlighted by sites such as LinkedIn,

in vocational progression (Buettner 2017). However, prior to the

introduction of social networking services (SNSs), gaining these

benefits typically required in-person meetings to develop a foun-

dation of trust. In the digital age, electronic invitations link secon-

dary and tertiary connections with no direct contact. This lack of

intimate knowledge of the person undermines the quality of refer-

ence, although Davis’s(Davis et al. 2020) examination of LinkedIn

usage found that online networking clearly generated career bene-

fits and supported use of career-oriented social networking markets

(CSNMs) as a prominent factor in career self-management success.

Two types of interconnections emerge in social capital theory,

bonding, where strong ties form within closer organizations, and

bridging, which uses weak ties to reach beyond the group (Alaa

et al. 2018;Neumeyer et al. 2019). Forms of bridging capital in-

clude nonoffice connections highlighted by school, alumni, and

professional organizations. Leveraging these channels provides ac-

cess to unique information. In the male-dominated AEC industry,

homophily persists, reinforcing leadership’s bonding social capital

and subsequent hiring of the next generation (Alaa et al. 2018).

Faced with altering the status quo, as women enter management,

they tend to expand their focus to incorporate bridging social

capital, recognizing the added benefits of larger networks in their

continued growth (G´orska et al. 2022).

Conversely, individuals without effective support systems lack

beneficial connections, delaying progression into leadership posi-

tions (Roebuck and Smith 2013). Early and midcareer workers pur-

suing career growth should strive to amass quality contacts to

recognize the benefits. Notably, two studies found that CSNMs

of 150 and 157 connections, respectively, offer the optimal size

to locate new jobs, whereas larger groups produce lower success

rates (Buettner 2017). Recruiters often view excessively large net-

works as noisy information, suggesting weak connections through-

out the group (Buettner 2017). This contrasts the perspective that

larger networks enhance women’s skills, including compliance,

civic virtue, and organizational development (Roebuck and Smith

2013), and that weak ties contribute to optimizing social capital

(Seibert et al. 2001). Most effective implementation of SNS net-

works lies not in the sheer size and volume of connections but

the understanding of how those individuals can positively impact

their career (Davis et al. 2020). Ultimately, balancing a larger group

with one’s ability to genuinely connect dictates the value of a

network.

References with knowledge of a candidate’s past professional

experience often generate prospects for career growth, whether

within the current company or a move to an external opening.

The expansion of the Internet over the last 25 years and associated

emergence of social media facilitated exponential connections with

local and remote colleagues. Launched in 2003, LinkedIn entered

the professional networking space and grew into an industry leader.

Considering women create larger formal and informal career net-

works than men (Roebuck and Smith 2013), would results reflect a

disparity in size?

J. Manage. Eng., 2022, 38(6): 04022060

H2: Faced with the disadvantage created by bonding social

capital homophily in the male-dominated AEC field, women seek

unique knowledge obtained by assembling larger external bridging

social capital networks.

Social Cognitive Career Theory

Social cognitive career theory creates a framework to study career

development through the lens of educational and professional pur-

suits (Lent et al. 1994;Lent and Brown 2019). Its five interconnected

models, interest development, decision-making, performance in

education and vocational domains, satisfaction and well-being in

educational and vocational contexts, and career self-management,

offer a comprehensive analysis of career choice motivations and out-

comes (Lent and Brown 2019). Even before entering the workforce,

SCCT self-efficacy drives students to select academic fields aligned

with their personal interests, leading to early career choices (Fouad

and Santana 2017). During midcareer, when considering additional

formal education or professional training, workers balance these in-

ternal interests with external professional influences. Ultimately,

SCCT creates a structure linking three interrelated aspects of career

development: career interests, academic and professional choices,

and their performance in these pursuits (Lent and Brown 2019).

Its principles provide a path to understand the factors creating

continued women’s underrepresentation (Blaique and Pinnington

2021).

Twenty-first-century careers embody mobile, nonlinear paths,

changing directions in response to both personal and professional

influences, including temporarily opting out for family and other

interruptions (Knowles and Mainiero 2021). SCCT provides a con-

textual reference on how self-efficacy beliefs, outcome expecta-

tions, and personal goals affect educational and career decisions

(Burga et al. 2020). Selection of career choices aligns with one’s

interests and is based in part on self-confidence (Lent et al. 1994).

Further, SCCT offers perspective to evaluate one’s career progres-

sion and analyze prior choices and subsequent outcomes (Fouad

and Santana 2017). Among the variable factors, Lent’s theory as-

sessed the impact of gender and culture on career advancement

(Lent and Brown 2019). SCCT’s framework shows that perceived

sexism builds a barrier for women and impacts their decisions

throughout a career (Fouad and Santana 2017). As a result, women

felt the need to repeatedly prove themselves capable to work along-

side their male counterparts and exert the extra effort to prepare

themselves for the same work (Clayton 2022). This self-perpetuating

cycle drives women out of the workforce or to other firms as they

believe advancement opportunities are unavailable (Knowles and

Mainiero 2021). Findings from this LinkedIn study confirm this

external pressure generates higher turnover frequency for women

and alters their career trajectory. Even those who temporarily opted

out of the workforce often decide to find new firms rather then return

to their former employer due to preexisting discriminatory practices

(Knowles and Mainiero 2021).

Primary aspects of SCCT include self-efficacy, social supports,

and goal setting, which shape one’s educational and professional

journey throughout their decades-long career (Lent and Brown

2019). Prior SCCT studies connected perceived entry barriers

and the resultant lack of participation by segments of the popula-

tion, recommending an industry response of altering recruitment

and retention strategies to increase diverse representation (Clayton

2022). Firms’reinforcement that employees and managers possess

the correct skills and abilities, coupled with the belief that their ed-

ucation and experience align with industry goals, increases the

quality of work and likelihood of retention (Burga et al. 2020).

Further, SCCT suggests that success in past projects reinforces

self-efficacy beliefs and generates better future performance (Burga

et al. 2020).

Number of Career Positions

Women and men experience different paths climbing the corporate

ladder (Lyness and Thompson 2000). Guy (2020) discussed the

“broken rung”impeding progression into entry-level management

positions, delaying women’s long-term progression into executive

billets. Assessing the current industry leadership teams’LinkedIn

profiles, would women require more total jobs to reach higher rungs

on the ladder? Whether remaining with a single company or work-

ing for multiple firms, how many rungs exist on the climb up the

corporate ladder? Would women face additional long-term chal-

lenges from smaller, incremental steps compared to male counter-

parts benefiting from fewer, larger career leaps? Regardless of

number of firms, collected data identifies each reported title. When

a candidate switched companies and retained the same title [e.g., vice

president (VP) to VP], these constitute two positions.

Women engineers build knowledge and skill through experience

and professional development in preparation for career advance-

ment (Marinelli et al. 2021). Blaique and Pinnington (2021) pro-

posed implementing the career self-management model (CSMM),

an extension of SCCT, to help women develop adaptive behaviors

to respond to negative career events. These adverse occurrences, in-

cluding lower promotion potential, sexist culture, and hostile work

environment, deteriorate the greater drive women possess at the be-

ginning of their careers than men, discouraging them and draining

their motivation (Knowles and Mainiero 2021). Human resources

departments creating a CSMM incorporating the needs of both

women and men helps them commit to their occupations and retain

their positions despite negative events (Blaique and Pinnington

2021). Regardless, women experience higher representation in non-

engineering occupations compared to their percentage of overall

engineering graduates (National Academy of Engineering 2018),

suggesting they either leave or never enter engineering at higher

rates compared to men.

H3: Adverse career occurrences disproportionally impact

women, slowing career progression and creating a greater number

of positions, including company changes, required to reach

leadership.

Highest Level of Education

This study considered impact of accumulated education, specifically

graduate studies, on women’s and men’s respective career arcs.

A key cornerstone of SCCT, education selection significantly im-

pacts a career arc (Lent and Brown 2019). Would women achieve

the same level of organizational success with comparable education?

Since the mid-1980s, women have entered the workforce at the same

rate as men but remain significantly underrepresented in leadership,

even for graduates of leading institutions (Ely et al. 2011). By the

early 2000s, women’s college graduation rates exceeded those of

men (Jardina and Burns 2016;Sandberg 2013), and women remain

well represented at lower and middle management positions but not

in the executive suite (Roebuck and Smith 2013). Despite more for-

mal education, one study found that women received significantly

lower levels of craft-specific training (Perrenoud et al. 2020), a po-

tential contributor to slower advancement.

Despite gains over the last few decades, selection of college

majors often aligns with self-perceived femininity or masculinity

(Beutel et al. 2018). Traditional roles reinforced by society contrib-

ute to precareer “leaky pipe”drain by syphoning off quality talent

even before reaching the workforce. Despite these hindrances,

J. Manage. Eng., 2022, 38(6): 04022060

women earned 21.9% of US civil engineering degrees conferred

between 1995 and 2018 (Department of Education 1999;Yoder

2019).

Although this study concentrated on engineering, career pro-

gression often shifts between functional disciplines, especially as

management positions require auxiliary soft skills beyond hard

technical engineering knowledge. As such, earning an undergradu-

ate engineering degree may not directly lead to an engineering career

or represent only the start of a career. In Marinelli et al.’s(2021)

study of engineers’progression into management and leadership,

women felt the need to reestablish their engineering competence and

credentials throughout their careers. This repeated search for self-

efficacy follows the SCCT model. Additional advanced degrees, es-

pecially for executives, include business management, finance, and

administration. One prior study found that only 36% of engineering

baccalaureates and 42% of engineering master’s degrees worked in

engineering positions. Contributing to this divergence, over 25% of

professionals with undergraduate engineering accreditation reported

managing people and teams as their primary responsibility (National

Academy of Engineering 2018). SCCT’s foundational focus on per-

sistence in educational domain progresses throughout a career (Lent

and Brown 2019). Alternatively, nonengineering undergraduates

working in the AEC industry could pursue graduate engineering de-

grees to align with their career trajectory in the industry.

In today’s workforce, across most US industries, graduation

from a traditional 4-year college or university typically serves as

a prerequisite, especially for management positions. This study

quantifies earned education, focusing on the highest level achieved.

Although not mandatory, graduate degrees further differentiate can-

didates and aid their ascension into executive leadership. Compar-

ing women and men, would the level of education be comparable

for similar positions, or would women need to accumulate addi-

tional schooling to maintain pace?

H4: Women continually reevaluate their preparedness and

performance in educational domains, attaining additional advanced

degrees to achieve similar organizational hierarchal status as men.

Research Methodology and Data

Data Collection and Cleaning

With women holding 10.9% of the total positions (US Bureau of

Labor Statistics 2020), AEC remains a malecentric field. An exten-

sion from prior research, this analysis focused on individual and

collective career histories of 2,857 industry leaders extracted from

LinkedIn profiles. Reviewing leading firms’websites listed on the

2019 Engineering News Record (ENR) Top 400 list, our research

team assembled the list of leadership teams publicly reported by

individual companies. Firms’classification of their respective se-

nior and executive managers impacted the available sample.

With the explosion of technology and relatively recent introduc-

tion of SNS, available online data significantly increased. The ap-

plication of web scraping automatically extracts large quantities of

information from web-based sources (Patel 2020). This method

quickly creates structured data sets from unstructured content for

further analysis. After assembling leadership teams from websites,

often listed under “Leadership”or “Meet Our Team,”researchers

located the matching LinkedIn profiles. Initial validation data points

included names, position titles, and, if available, photos. Consider-

ing some industry firms, especially family-owned companies, pos-

sess multigenerational teams, reviews applied extra care to ensure

accuracy.

Permission to access all collected information required only

a basic LinkedIn account with no special connections or links. This

approach maintained equal access of data rights for all profiles.

Applied to unique URLs, Python script coded using the available

packages Urllib and BeautifulSoup performed web scraping, cap-

turing information including companies, titles, years of experience,

education, and network size. Web scraping facilitates repeatable

high-speed data extraction, ease of code adjustment, and structur-

ing for analysis. This flexibility enabled multiple rounds of review,

code adjustments, and data cleaning (Diouf et al. 2019).

Three rounds of data validation and cleaning maximized the

accuracy of the final data set. During the initial cycle, we tested

a smaller sample where two researchers independently manually

validated 20% of records. Combining results of identified discrep-

ancies, code adjustments improved the quality of the data pull. As

an additional check on the compiled data set, the authors assessed

whether differences existed between women and men in the level of

reporting for individual profiles. Would one gender provide fewer

details in their career history (e.g., omit early career positions, de-

gree earned, or years served)? Before conducting further analysis,

the team evaluated the data integrity, specifically analyzing each

educational listing as split into three individual components, school,

degree, and graduation year, to compare listed level of detail. No

statistical differences existed between women and men; therefore,

researchers determined consistent levels of accuracy existed, which

increased the confidence level about the data set.

Statistical Test and Analysis—Machine Learning Model

Development

ML refers to a set of algorithmic structures designed to provide

computer systems with the ability to learn and improve their per-

formance via the discovery of patterns in data. Applying ML algo-

rithms to engineering problems develops predictive models based

on a set of training features. The accuracy of such predictions gen-

erally depends on the amount and quality of input data, as well as

the algorithm’s ability to extract the patterns (Ray 2019). Generally,

practitioners consider ML models such as supervised, unsupervised,

and reinforcement learning techniques. Supervised ML involves user

specification of algorithm targets. Thus, algorithms focus on model-

ing the relationship between a set of input variables (e.g., career path

features) and the specified target (e.g., gender). Flexibility of ML

models in comparison to statistical models, which typically include

rigid requirements and the clear explanation of model’s accuracy us-

ing testing data set for validation, justify applying ML models in this

study. Therefore, the authors applied supervised learning classifiers,

cases with categorical targets, to investigate the impact of gender on

individual career paths as a whole picture.

Models easily learn patterns to classify gender when significant

different relationships exist in the career paths of female and male

construction leaders. Although many ML techniques exist in re-

search and each technique involves its own unique attributes, a

group of top-used ML classifiers apply in the construction studies

(Seyrfar et al. 2021). In this study, the researchers employed ran-

dom forest (RF) to investigate the relation of career path features

and target class of gender. A flexible, straightforward, and easy-to-

use ensemble ML algorithm, RF yields highly accurate results deal-

ing with various input data types. Merging multiple base decision

trees, RF produces a single optimal predictive model (Zhang and

Ma 2012). Further, parallelization contributes to reducing over-

fitting, overcoming bias issues, and enhancing classification per-

formance (Jiang et al. 2020). These features, in addition to the

capability to rank feature importance, offer a useful algorithm to

implement in this study.

J. Manage. Eng., 2022, 38(6): 04022060

In terms of application, ML classifiers demonstrate significant

contributions in a recent variety of related engineering and con-

struction issues. Jafari et al. (2020) developed a ML model to

classify potential employees under wage inequality in engineering

organizations. Awada et al. (2021) trained a RF classifier to pre-

dict acceptance or rejection of contractor-submitted inspection re-

quests. Khalef and El-adaway (2021) developed a ML classifier to

categorize substantial and nonsubstantial changes in construction

projects. Moreover, researchers applied ML classifiers to extract

textual information from social media posts (Li et al. 2021). This

study of LinkedIn profiles employed ML classifiers to categorize

female and male construction career paths.

Model development included data preprocessing, algorithm test-

ing, model validation, and feature importance extraction. Among

multiple cross-validation techniques, stratified random sampling

evaluated the performance of classifiers using the number of true

positive (TP), true negative (TN), false positive (FP), and false neg-

ative (FN) results based on predicted and actual classes. Then mod-

els calculated performance metrics: precision, recall, and F1 score

(Liberty et al. 2016).

Results and Discussion

The compiled data set segregates each individual reported company

to quantify movement throughout the individual’s career. Python

software packages Urllib and BeautifulSoup extracted LinkedIn

profiles and segmented individual candidates’listed company his-

tory. Coding cleaned the data set, identified and removed unmain-

tained profiles (e.g., a president with only one position listed for the

last 40 years), and counted the number of unique firms throughout

individual careers. Accumulated totals, sorted by gender and engi-

neering or nonengineering, formed the resultant data set for analy-

sis. This basic methodology applies to all research questions.

This study focused on executive and senior level leaders includ-

ing presidents, C-suite officers, executive vice presidents, VPs,

general managers, and directors. Further, analysis segregated engi-

neering and nonengineering billets to align with technical focus

of the field. The sample (Table 1) comprised 2,857 total profiles

(422 women and 2,435 men). Of the executives identified across

the ENR Top 400 firms, women filled 14.8% of the total positions,

including 5.2% of engineering and 33.2% of nonengineering

roles. Consistent with the prior study in this research series, dispro-

portional levels of finance, legal, and business development roles

necessitated this separation (Hickey and Cui 2020). Although these

values lag compared to the overall women’s percentage of the US,

these figures exceed women’s overall industry total of 10.9%

(US Bureau of Labor Statistics 2020).

Even though women earn a sufficient proportion of formal civil

engineering degrees (Department of Education 1999;Yoder 2019)

and form an industry-focused candidate pool, the number of females

holding engineering positions decreases with increasing organiza-

tional levels. Assessing progress of this study compared to the pre-

ceding 2019 research of this multiphase study (Hickey and Cui

2020), executive market penetration regressed from 3.9% to 3.8%.

The climb to the top remains daunting despite progress in other

areas, with continually fewer filled billets with each organizational

hierarchal step from director (8.8%) to VP (5.5%) and ultimately

executive suite (3.8%).

Predictive Machine Learning Analysis

Data preprocessing forms a critical step in developing ML models.

Many ML algorithms cannot operate on categorical data directly,

requiring all input variables in numeric format. Categorical varia-

bles with no such ordinal relationship apply one hot encoding to

transfer data into the appropriate format for further analysis (Seger

2018). Table 2presents the list of career path attributes extracted

from LinkedIn for each professional. In the first step, using the avail-

able Python packages such as Matplotlib and Seaborn, the authors

plotted the relation of career path features. Effective data visualiza-

tion transforms a complicated data set into an understandable for-

mat. Fig. 1includes multiple joint plots that describe the behavior of

career path features comparing gender classes. All available data

points generated each component of Fig. 1. Some profiles omit ed-

ucational background; therefore, the sample size in Fig. 1(e) differs

from other figures. According to Fig. 1(a), women’s distribution of

number career positions and companies differs from men leaders.

Fig. 1(b) suggests women need to change more companies to

achieve their current role. Fig. 1(c) demonstrates that females re-

ported a larger professional network overall. Fig. 1(d) illustrates

the importance of company size on leaders’career paths, specifically

showing that females generally work in larger companies. It explains

that large company leaders possess significantly more experience.

Finally, Fig. 1(e) shows that women leaders achieve higher educa-

tion degrees compared to their male peers. Data visualization

explained a general picture of potential differences. In the following

steps, we dig into how to consider multiple career path features in the

ML model and provide detailed quantitative analysis for each career

path feature.

Deploying a Python program using the Sklearn library, the re-

searchers built a pipeline to apply a ML classification algorithm.

ML algorithms tend to be biased towards classes with greater in-

stances, and imbalanced classes risk generating unsatisfactory clas-

sifiers. Therefore, the authors followed an undersampling approach

Table 2. Career path attributes

Variable Data type Description

Number of career firms Numerical Number of unique companies

Number of career positions Numerical Number of unique career positions

Highest level of education Categorical Highest achieved degree (e.g., Ph.D., MS, MBA, JD, BS)

Professional network size Numerical Number of followers

Years to current position Numerical Duration to achieve the current position

Company revenue size Categorical Under $250M, $250M–$500M, $500M–$1B, Over $1B

Position type Categorical Executive suite, vice president, director

Role type Categorical Engineering, nonengineering

Table 1. Study population composition

Class Engineering Nonengineering Total

Female 97 (5.2%) 325 (33.2%) 422 (14.8%)

Male 1,780 (94.8%) 655 (66.8%) 2,435 (85.2%)

Total 1,877 980 2,857

J. Manage. Eng., 2022, 38(6): 04022060

to balance classes distribution, including 10 repeated random

selection of male profiles, placing them beside all female profiles

and reporting average results (Haixiang et al. 2017). Following

multiple iterations, the balanced data set divided into 70% train-

ing and 30% testing samples. Building the RF model requires

important features established such as the number of trees

(n_estimators), maximum number of levels in each tree (max_

depth), maximum number of features considered for splitting at

each node (max_features), and method for resampling with or

without replacement (bootstrap). The process of calculating the

best values for the model’s parameters is called hyperparameter

tuning (Probst et al. 2019). Using the Sklearn library hyper-

parameter grid search, the authors optimized the RF model’s

parameters. The resultant optimized RF model reported high per-

formance with 98.95% training and 89.53% testing accuracy, in-

cluding a balanced performance for both classes, which offers a

positive indicator of the potential impact of gender on construc-

tion leaders’career paths. Table 3summarizes the RF model per-

formance indicators. Among the important RF model parameters,

selected values include n_estimators = 100, max_depth = 12,

max_features = Auto, and bootstrap = True.

Using the Sklearn library feature importance attribute, career

path elements were sorted based on their significance to explaining

the target gender class. Fig. 2outlines attribute importance in the

RF model, calculated by averaging the decrease in impurity over

trees (Han et al. 2016). In summary, model results reveal signifi-

cantly different patterns in the career paths of female construction

leaders compared to their male counterparts. In other words, view-

ing elements of leaders’career arcs, women follow a different path

to arrive at the top of the corporate ladder. Important features

Fig. 1. Data set career path feature visualization: (a) number of career companies; (b) years to current job; (c) network size; (d) company size; and

(e) highest level of education earned.

Table 3. Random forest detailed performance

Class Precision Recall F1 score Number of tests

Female 0.89 0.89 0.89 115

Male 0.90 0.90 0.90 125

Average 0.90 0.90 0.90 240 Fig. 2. Feature importance sorting.

J. Manage. Eng., 2022, 38(6): 04022060

classifying construction leaders’career paths, highlighted by com-

pany revenue size and role type, reinforce key variables found in

the prior study by the authors, including higher representation of

women in larger firms and nonengineering roles (Hickey and Cui

2020). Other important items sorted as number of career firms, pro-

fessional network size, and number of career positions. The follow-

ing sections detail statistical comparisons of important career path

components to evaluate the study’s proposed hypotheses.

Statistical Comparison of Career Path Features

Number of Career Firms

Comparing careers of women and men finds that women switch

firms 56% more frequently, 4.2 to 2.7 companies, during their ca-

reers (P<0.001). A key finding centers around leaders remaining

with a single firm throughout their careers. Prior research found that

internal candidates ascend faster than external ones (Koch et al.

2017). That study, however, reports that faster promotion for internal

candidates applies more to men than women. These strong ties,

bonding social capital (Neumeyer et al. 2019), serve to continue the

male-dominated hierarchal structures. As shown in Fig. 3, almost 4

in 10 men remain with one company (37%), whereas that figure

drops to 10% for women. Extrapolating to early and midcareer pro-

fessionals, internal vertical progression benefits men, whereas their

female counterparts should consider moving to another company.

Social capital theory studies found that colleagues’similarities, in-

cluding sex, age, and ethnic and racial backgrounds, contributed

to strengthening network linkages, whereas greater differences cre-

ated a higher likelihood of organizational turnover (Brass 2003).

Low numbers of women in executive billets contributes to differen-

ces and greater frequency of turnover for women overall.

H1 (True): Male-dominated executive teams’bonding social

capital perpetuate the status quo. Study results demonstrate 56%

higher turnover for women than men during their career progression,

4.2 to 2.7 companies. Fifty-eight percent of men and only 27% of

women achieve leadership positions with only one or two firms.

Professional Network Size

Women demonstrate superior communication skills, develop better

connections with stakeholders (Devnew et al. 2018), and cultivate

larger formal and informal career networks than men (Roebuck

and Smith 2013). A gender-based study of engineers and their

continued pursuit of knowledge found that women perceive the

highest accessibility of information from other women, whereas

men seeking knowledge from women are the lowest (Poleacovschi

et al. 2021). Analyzing the size of assembled networks, results sup-

port women’s ability to create more connections. Women generate

14% larger groups (1,200 to 1,049) than men (P¼0.011) (Fig. 4).

In social capital theory, larger networks create a greater volume of

weak ties and structural holes (Seibert et al. 2001). This bridging

offers not only access to unique data but also the opportunity to

control access to the acquired information (Alaa et al. 2018), which

could explain the higher turnover rate and number of firms.

More experienced women build stronger networks, suggesting

that mentoring would benefit early or midcareer professionals

(Neumeyer et al. 2019). Cultivating professional contacts early and

often maximizes their growth potential. However, Buettner (2017)

argued that an individual can only maintain 100–200 stable rela-

tionships with a decreasing value of larger networks. Coworkers,

clients, vendors, and competitors constitute LinkedIn connections,

which include passing introductions at business meetings, industry

conferences, and opportunistic “professional networkers.”Reflect-

ing on the sheer size of the networks, how can an individual main-

tain valuable interactions? Although averaging 1,200, 59% of

sampled women assembled followers under 1,000. Extreme results

return 32 profiles listing 200 or fewer connections and 33 with

over 3,000.

An item of note in the data collection, due to LinkedIn privacy

settings beyond first connections, connection counts above 500 are

displayed as “500+.”In order to retrieve consistent information for

all candidates, the authors would need to connect directly with each

of the 2,857 leaders in this survey. Recognizing the challenges of

obtaining these linkages, authors chose to quantify network size

through use of the followers, which includes an uncapped limit.

H2 (True): In order to offset men’s bonding social capital ho-

mophily, women seek unique knowledge obtained by assembling

14% larger professional networks, 1,200 to 1,049, than men.

Additionally, a sign of interactivity and strength of relationships

manifests itself in professional recommendations. Women actively

assist fellow professionals by offering and receiving 64% more rec-

ommendations (2.1 to 1.3) (P<0.001). In the sample, less than half

of either gender submitted or received recommendations. Of those

demonstrating activity with over five total, women outpaced men

by a factor of two, 14% to 7%.

Fig. 3. Number of career companies. Fig. 4. Network size.

J. Manage. Eng., 2022, 38(6): 04022060

Number of Career Positions

Career mobility itself embodies different paths. Progression may

involve promotion through the technical job family tree (e.g., engi-

neer, senior engineer, engineering manager, and beyond into lead-

ership), changing between functional areas (e.g., engineering to

finance), changing employers, or a combination. Aside from the

possibility of vertical functional paths, the remaining options often

occur when employees perceive superior opportunities exist or

they lose their current position. “Climbing the corporate ladder”

embodies a common path to the executive suite. SCCT requires

continual reassessment of career decisions (Lent et al. 1994;Lent

and Brown 2019), including impact of negative outcomes (Blaique

and Pinnington 2021). The authors assessed whether women and

men climbed an equivalent number of “rungs”to reach leadership.

On average, women held almost one more position, 5.8 jobs to 4.9,

19% more (P<0.001) (Fig. 5).

As outlined in the “Data Collection and Cleaning”subsection,

before conducting the data analysis, the authors assessed whether

women and men provided similar detail in reporting career accom-

plishments. Using three components of educational data: school,

degree, and graduation year, no statistical reporting differences

were identified between genders, increasing the confidence level

of data integrity. As shown in Fig. 5, men consistently listed fewer

positions to reach leadership, indicating that women needed to

climb additional rungs on the corporate ladder to achieve similar

positions. Twenty-two percent of women and 37% of men reported

two or three career positions, whereas at the other extreme, women

listed 10 or more positions twice as often as men (12% to 6%).

H3 (True): Adverse career occurrences disproportionally im-

pact women, slowing career progression and requiring 19% more

positions, 5.8 compared to 4.9, on the corporate ladder than their

male counterparts during their career progression to leadership.

Reviewing more deeply to assess any impacts of company size,

regardless of revenue levels,women consistently held more positions

(Fig. 6). However, both genders reported increasingly frequent title

changes with corresponding organizational magnitude. Larger firms

typically integrate additional hierarchical layers of management

because the employee base of a $250M company allows a flatter

organization than one over $1B. This creates extra rungs, driving

higher incidence of job changes to reach leadership. Although mem-

bers’mobility includes shifts between small and large firms, Fig. 3

shows the propensity toward less frequent changes for men.

Highest Level of Education

Analyzing educational histories for the sample (Fig. 7), more than

half of women earned advanced degrees (53.9%), whereas men

achieved these milestones less 1=3of the time (31.2%). These re-

sults support earlier studies, which found women earn the majority

of all college degrees (Jardina and Burns 2016;Sandberg 2013).

Subdividing Fig. 7results by type of graduate degrees, Table 4

offers a more detailed insight into how both genders focused on

advanced studies, specifically the directions that women pursue to

further their careers. As previously discussed, most positions filled

by women are classified as nonengineering. These include legal

Fig. 5. Number of career positions.

Fig. 7. Highest education level.

Table 4. Breakdown of graduate degrees

Degree Female (%) Male (%)

Ph.D. 1.7 0.5

JD 12.2 4.1

MBA 20.0 16.4

MS/MA 20.0 10.2

Total 53.9 31.2

Fig. 6. Average number of positions held by company revenue.

J. Manage. Eng., 2022, 38(6): 04022060

roles, supporting the disparate volume of juris doctor (JD) degrees,

12.2% for women compared to only 4.1% of men.

Pursuit of and satisfaction with one’s continued education

forms a cornerstone of SCCT in the analysis of a career trajectory

(Lent and Brown 2019). This focuses on interaction between self-

efficacy and expected outcomes, positively impacted by earning

graduate degrees (Clayton 2022). Further, best practices for for-

warding a career include formal education, training, and mentor-

ing (Schultheiss 2021).

H4 (True): Women continually reevaluate their preparedness

and performance in educational domains, obtaining more advanced

degrees, 53.9% compared to 31.2%, than men to achieve similar

organizational hierarchal status as men.

Years to Current Position

Reflecting on initial study results, specifically that women hold

more positions on the corporate ladder, researchers considered

whether the time to reach their current position required additional

years of service. Analysis calculations considered the duration from

the initial reported employment to the start of the current job

(e.g., a professional life started in 1995 and holding a VP title from

2015 to 2021 equals 20 years, not 26). Although not statistically

significant (P¼0.242), actual results returned that women

achieved their current position with the same accumulated years

of experience.

Managerial Implications

Practical Applications

AEC firms miss out on the benefits derived from assembling teams

possessing varying backgrounds and experiences. Increasing diver-

sity in the industry injects new perspectives when assessing busi-

ness opportunities and problem solving into the daily workplace.

However, this represents only a secondary application offered by

this study. The authors primarily focused on collecting and analyz-

ing career paths of successful AEC women and men to assess

whether discernable patterns existed in behaviors and decisions

throughout the decades-long evolution of individual careers. Sub-

sequent cumulative results showed distinctly different paths be-

tween genders from the LinkedIn profiles of 422 women and

2,435 men. The results offer potential roadmaps for early and mid-

career professionals to assess when considering key decisions. In

addition to a primary choice of remaining with an organization or

changing companies, pursuit of further education and assembling

professional networks impact growth opportunities. Specifically,

results suggest that women should strive to earn more advanced

degrees, create larger networks, and seek opportunities to change

firms more often than men. Developing policies and programs sup-

porting these areas assist firms to attract and retain top talent.

Company Impact

Effectively implemented, gender diversity creates an environment

providing women and men fair and equal treatment. Unfortunately,

women perceive greater obstacles in joining and advancing in AEC

(Infante-Perea et al. 2021). Firms lacking healthy inclusive cultures

generate higher turnover of women leaders (Fig. 3). This prevents

advancement of qualified people based solely on gender, including

driving them out of the pool (True-Funk et al. 2021) and limiting

overall participation of women, currently 5.2% of engineering lead-

ership billets (Table 1). Whereas the origins of affirmative action

focused on gaining entry to the market, even 30 years ago, advance-

ments beyond entry-level positions stalled at middle management

(Thomas 1990). This creates poor human resources practice, not

only for short-term organizational performance but also long-term

signals to potential future leadership candidates (Powell 2020).

Firms must identify underlying reasons for higher turnover and de-

velop recruiting and retention techniques and practices that address

the root causes in order to retain this talent. Hindering progress,

Table 1shows the current dearth of women in leadership roles,

offering few role models for current women workers and middle

managers. Despite AEC industry efforts to attract and retain more

women, even those who reach supervisory roles leave their posi-

tions more often than men (Maurer et al. 2021). This detracts from

the ability to guide younger people into future leaders. Firms

instituting formal and informal networking programs increase

cross-functional connections and foster a more cohesive working

environment. Educating the team on the role of social capital levels

the playing field for everyone (G´orska et al. 2022). A key challenge

of successfully implementing a diversity program remains identi-

fying and training qualified candidates, not just increasing the

numerical volume for optics (Thomas 1990). Will firms integrate

proper identification, mentoring, and training programs to increase

internal qualifications and the subsequent talent pool? Anecdotally,

what is worse than investing in an individual, and they leave? Not

investing, and they stay.

Education reimbursement offers companies a key element of a

competitive total compensation package and demonstrates a com-

mitment to long-term employee growth. As shown in Fig. 7and

Table 4, women earn advanced degrees at a higher rate than their

male counterparts, 53.9% to 31.2%. Understanding the expenses

associated with implementation of continuing education benefits,

firms should evaluate costs related to employee turnover and tailor

programs to ensure both women and men benefit, especially for

engineering-related disciplines.

As a sign of changing times and perspectives, company focus on

diversity evolved from checking a box to an integral component of

culture and financial performance (Morgan 2020). On the other

hand, premature departure of women (Fig. 3) drains unrealized po-

tential gains. Firms prioritizing retention and growth programs po-

sition themselves with a competitive advantage (Hewlett and Luce

2005). With the growing societal discussion of diversity and fair-

ness, propelled by the accelerating speed of social media, company

actions create greater risk or reward results. Establishment of a

thoughtfully crafted diversity program promises positive short-term

and long-term benefits. Ultimately, companies need to reenvision

how they see their workforce and humanize the culture.

Early and Midcareer Professionals

Commonly recognized as a beneficial career-enhancing activity,

mentoring provides practical advice and recommendations directly

from senior professionals. Executives pass along experiences, pos-

itive and negative milestones, and lessons learned gained over a

career to aid the next generation’s growth. In its best form, mentor-

ing imparts knowledge that allows the protégé to capitalize on

career-enhancing activities and avoid pitfalls of their predecessors.

Compilation of this LinkedIn corpus of career decisions gathers ex-

periences of 2,857 successful women and men for the benefit of

early or midcareer professionals. Equal availability of the results,

especially for those without access to personal mentoring or margin-

alized by homophily, partially levels the imbalance of opportunities.

This article identifies actions taken by senior and executive

women and men in their individual and collective career journeys

to achieve success. Accompanying recommended actions (e.g., con-

sider switching companies) reflects these prior paths. The authors

wish to note that changing the paradigm to expand opportunities for

J. Manage. Eng., 2022, 38(6): 04022060

women resides with AEC companies and intend no implication that

women must shoulder the responsibility or created the current

imbalance.

Prior research found that women navigate different paths to ex-

ecutive positions than men (Lyness and Thompson 2000). Through

an analysis of over 2,800 executives’careers, this study confirmed

identifiable career patterns based on gender. Although each individ-

ual encounters unique opportunities and challenges, results offer

suggestive paths for women to navigate the male dominated exec-

utive levels.

Women experience higher turnover rates, with only 27% of en-

gineering leaders remaining with one or two companies during their

career. Conversely, men ascend at more than double that rate, 58%,

for the sample. Similarly, women reported three or fewer positions

22% of the time compared to 37% for men. Despite being counter to

evidence of an accelerated promotion path by remaining with the

same firm (Koch 2017; Hamori and Kakarika 2009), women seek-

ing advancement should carefully consider seeking opportunities in

other firms and areas. Further, a key element of SCCT, accumulation

of additional training and education, aligns with career progression.

Results show the importance of advanced degrees for females to

maintain pace with their male counterparts. Despite over half of

the group, 53.9% (Fig. 7and Table 4) earning postbaccalaureate ed-

ucation, their progression still lags. Lack of documented knowledge

risks further career drag, so early and midcareer women should

consider pursuing continuing formal education. In the AEC field,

engineering undergrads possess the option of augmenting technical

knowledge with business and law degrees or continuing on the

engineering track.

Careful selection of mentors offers enhanced opportunities for

women to successfully progress. Prior research discovered that

mentored women reach middle management billets 56% more

often (Roebuck and Smith 2013). The results of this study demon-

strate that women executives understand the value of developing

relationships through the creation of 14% larger networks (Fig. 4).

Ascending in the corporate structure, current junior and midlevel

managers should consider leveraging their social capital skills,

creating larger networks, seeking out mentorship, and gaining a

high-level sponsor in the organization.

Conclusion

This study contributes to the body of knowledge by identifying ca-

reer choices by current women and men in the AEC industry.

Whereas prior studies discussed theories and offered insight into

origins and reasons behind the lack of diversity in AEC and other

fields, the authors found minimal information on career paths of

senior and executive leaders, especially divided and compared

by gender. Application of ML techniques to validate observational

analysis confirmed women follow different paths than men in their

ascension on the corporate ladder.

Extracting executive profiles of 2,857 representatives in the

field, women move 56% more often, require 19% more additional

jobs, earn more education (53.9% to 31.2% graduate degrees), and

leverage social capital to grow 14% greater professional networks

yet only fill 14.8% of the total executive pool, declining to 5.2% of

engineering leadership positions. Testing the RF classifier from the

assembled data set demonstrates the significance impact of gender

on career path data patterns by achieving accuracy of 98.95% dur-

ing training and 89.53% during testing to predict the gender of a

profile based on the career path features.

Attracting and retaining top talent requires inclusion of the wider

workforce. Unfortunately, women experience greater obstacles in

joining and advancing in AEC. Peering into the future, current lack

of prominent women in the workforce threatens to hinder expansion

of the available pool of candidates to create diverse teams and rec-

ognize the associated benefits. Reversing underrepresentation of

women in the industry requires comprehensive and collaborative ef-

forts from industry, academic institutions, and policymakers. Organ-

izations should examine their policies and practices for hiring,

promotions, and compensation benefits to ensure fair treatment of

all employees.

Limitations of this study include accuracy and consistency of

self-reported data level of detail. Although the authors considered

the public nature of the information and specter of potential neg-

ative professional ramifications the equivalent of the data being

professionally peer reviewed, specific experiences and education

may be underreported or overreported, especially for candidates

with longer careers and higher volume of companies and positions.

Further, with no direct contact with individuals, researchers applied

existing concepts, including social capital theory and SCCT, to data

trends. Further investigation into internal and external influences

would offer better insight into career decisions. Additionally,

AEC industry would benefit from research into other industries’

initiatives to improve diversity and investigate refinement and ap-

plication in the field.

This study represents the second phase of AEC gender diversity

research, building on prior quantification of executive industry rep-

resentation. Delving deeper into executive LinkedIn profiles, this

paper links to the planned third segment, which entails interviewing

key leaders and exploring their individual journeys. Applying foun-

dational data collected during this LinkedIn analysis, discussions

will focus on impact of job and company mobility, education,

and network development on their career progression. In addition,

external factors identified during the initial study, highlighted by

gender bias, sexism, and work–life balance, will be explored to

understand executives’resultant actions. Collectively, the authors’

intent focuses on expanding the pathway for women to advance in

the AEC industry.

Data Availability Statement

All data were collected through publicly available information

sources. Raw data and data profiles will be available from the cor-

responding author by reasonable request subject to the LinkedIn

data sharing and privacy policies.

Acknowledgments

Ms. Veronica Hercules and Mr. Adam Johnson from the University

of Maryland helped with data collection and cleaning. Their efforts

proved invaluable to the construction of this paper. Comments from

Journal of Management in Engineering reviewers improved the the-

oretical underpinnings, hypothesis refinement, and overall structure

of the manuscript.

References

Aisenbrey, S., and A. Fasang. 2017. “The interplay of work and family

trajectories over the life course: Germany and the united states in com-

parison.”Am. J. Sociol. 122 (5): 1448–1484. https://doi.org/10.1086

/691128.

Alaa, A. M., K. Ahuja, and M. Van Der Schaar. 2018. “A micro-foundation

of social capital in evolving social networks.”IEEE Trans. Network Sci.

Eng. 5 (1): 14–31. https://doi.org/10.1109/TNSE.2017.2729782.

J. Manage. Eng., 2022, 38(6): 04022060

Awada, M., F. J. Srour, and I. M. Srour. 2021. “Data-driven machine learning

approach to integrate field submittals in project scheduling.”J. Manage.

Eng. 37 (1): 04020104. https://doi.org/10.1061/(ASCE)ME.1943-5479

.0000873.

Baker, M., M. Ali, and E. French. 2021. “Leadership diversity and its

influence on equality initiatives and performance: Insights for construc-

tion management.”J. Constr. Eng. Manage. 147 (10). https://doi.org/10

.1061/(ASCE)CO.1943-7862.0002147.

Beutel, A. M., S. W. Burge, and B. A. Borden. 2018. “Femininity and

choice of college major.”Gender Issues 35 (2): 113–136. https://doi

.org/10.1007/s12147-017-9195-8.

Blaique, L., and A. H. Pinnington. 2021. “Occupational commitment of

women working in set: The impact of coping self-efficacy and mentor-

ing.”Hum. Resour. Manage. J. 1–45. https://doi.org/10.1111/1748

-8583.12405.

Brammer, S., A. Millington, and S. Pavelin. 2009. “Corporate reputation

and women on the board.”Br. J. Manage. 20 (1): 17–29. https://doi.org

/10.1111/j.1467-8551.2008.00600.x.

Brass, D. J. 2003. A social network perspective on human resources man-

agement. Networks in the knowledge economy, 283–323. New York:

Oxford University Press.

Buettner, R. 2017. “Getting a job via career-oriented social networking

markets: The weakness of too many ties.”Electron. Markets: Int. J.

Networked Bus. 27 (4): 371–385. https://doi.org/10.1007/s12525-017

-0248-3.

Burga, R., J. Leblanc, and D. Rezania. 2020. “Exploring student percep-

tions of their readiness for project work: Utilizing social cognitive

career theory.”Project Manage. J. 51 (2): 154–164. https://doi.org/10

.1177/8756972819896697.

Byrd, M. Y., and C. L. Scott. 2014. Diversity in the workforce: Current

issues and emerging trends. New York: Routledge.

Clayton, E. J. 2022. “The underrepresentation of women in building trades

programs and careers: Perceptions behind the educational and career

decision-making process.”Doctoral dissertation, Educational Leader-

ship Dept., Abilene Christian Univ.

Davis, J., H.-G. Wolff, M. L. Forret, and S. E. Sullivan. 2020. “Networking

via LinkedIn: An examination of usage and career benefits.”J. Voca-

tional Behav. 118 (Apr): 103396. https://doi.org/10.1016/j.jvb.2020

.103396.

Department of Education. 1999. Bachelor’s, master’s, and doctor’s

degrees conferred by degree-granting institutions, by sex of student

and field of study 1994–1999. Washington, DC: Institute of Education

Services, National Center for Education Statistics, Digest of Education

Statistics.

Devnew, L., M. J. Le Ber, M. Torchia, and R. Burke. 2018. “Introduction.”

In More women on boards: An international perspective, edited by

L. Devnew, M. Le Ber, M. Torchia, and R. J. Burke. Charlotte, NC:

Information Age Publishing.

Diouf, R., E. N. Sarr, O. Sall, B. Birregah, M. Bousso, and S. N. Mbaye.

2019. “Web scraping: State-of-the-art and areas of application.”In

Proc., 2019 IEEE Int. Conf. on Big Data. Los Angeles, CA: IEEE.

https://doi.org/10.1109/BigData47090.2019.9005594.

Dzombak, R., and K. Mehta. 2017. “Engaging and retaining women in

an engineering and entrepreneurship program [commentary].”IEEE

Technol. Soc. Mag. 36 (4): 15–52. https://doi.org/10.1109/MTS.2017

.2763452.

Eason, J., and B. Krause. 2001. “Networking.”IEEE Antennas Propag.

Mag. 43 (4): 138–141. https://doi.org/10.1109/74.951568.

Elliott, J. R., and R. A. Smith. 2004. “Race, gender, and workplace

power.”Am. Sociol. Rev. 69 (3): 365–386. https://doi.org/10.1177

/000312240406900303.

Ely, R., H. Ibarra, and D. Kolb. 2011. “Taking gender into account: Theory

and design for women’s leadership development programs.”Acad.

Manage. Learn. Educ. 10 (3): 474–493. https://doi.org/10.5465/amle

.2010.0046.

Fouad, N. A., and M. C. Santana. 2017. “SCCT and underrepresented pop-

ulations in STEM fields: Moving the needle.”J. Career Assess. 25 (1):

24–39. https://doi.org/10.1177/1069072716658324.

Frederiksen, A. 2008. “Gender differences in job separation rates and

employment stability: New evidence from employer-employee data.”

Labour Econ. 15 (5): 915–937. https://doi.org/10.1016/j.labeco.2007

.07.010.

G´orska, A. M., D. Dobija, Z. Staniszewska, and K. Prystupa-Rządca. 2022.

“Women’s journey to the top: The role of bonding and bridging social

capital.”Gender Manage.: Int. J. 37 (1): 77–93. https://doi.org/10.1108

/GM-01-2021-0004.

Grant, B. Z., and B. H. Kleiner. 1997. “Managing diversity in the work-

place.”Equal Opportunities Int. 16 (3): 26–32. https://doi.org/10.1108

/eb010685.

Guy, S. 2020. “Fixing the ‘broken rung’in the ladder to success.”In

Preparetopracticecuriosity, 42. Chicago: Society of Women

Engineers.

Haixiang, G., L. Yijing, J. Shang, G. Mingyun, H. Yuanyue, and G. Bing.

2017. “Learning from class-imbalanced data: Review of methods and

applications.”Expert Syst. Appl. 73 (May): 220–239. https://doi.org/10

.1016/j.eswa.2016.12.035.

Hamori, M., and M. Kakarika. 2009. “External labor market strategy and

career success: CEO careers in Europe and the united states.”Hum.

Resour. Manage. 48 (3): 355–378. https://doi.org/10.1002/hrm.20285.

Han, H., X. Guo, and H. Yu. 2016. “Variable selection using mean decrease

accuracy and mean decrease Gini based on random forest.”In Proc.,

2016 7th IEEE Int. Conf. on Software Engineering and Service Science

(ICSESS), 219–224. New York: IEEE.

Hasan, A., A. Ghosh, M. N. Mahmood, and M. J. Thaheem. 2021.

“Scientometric review of the twenty-first century research on women

in construction.”J. Manage. Eng. 37 (3): 04021004. https://doi.org/10

.1061/(ASCE)ME.1943-5479.0000887.

Hewlett, S. A., and C. B. Luce. 2005. “Off-ramps and on-ramps: Keeping

talented women on the road to success.”Harvard Bus. Rev. 83 (3):

43–46. https://doi.org/10.1108/hrmid.2008.04416bae.003.

Hickey, P. J., and Q. Cui. 2020. “Gender diversity in us construction

industry leaders.”J. Manage. Eng. 36 (5): 04020069. https://doi.org/10

.1061/(ASCE)ME.1943-5479.0000838.

Infante-Perea, M., E. Navarro-Astor, and R.-O. Marisa. 2021. “Sex, age,

work experience, and relatives in building engineering career develop-

ment.”J. Manage. Eng. 37 (5): 1–5. https://doi.org/10.1061/(ASCE)

ME.1943-5479.0000935.

Jafari, A., B. Rouhanizadeh, S. Kermanshachi, and M. Murrieum. 2020.

“Predictive analytics approach to evaluate wage inequality in engineer-

ing organizations.”J. Manage. Eng. 36 (6): 04020072. https://doi.org

/10.1061/(ASCE)ME.1943-5479.0000841.

Jardina, A., and N. Burns. 2016. “Advances and ambivalence: The conse-

quences of women’s educational and workforce changes for women’s

political participation in the United States, 1952 to 2012.”RSF: Russell

Sage Found. J. Social Sci. 2 (4): 272–301. https://doi.org/10.7758/rsf

.2016.2.4.10.

Jiang, T., J. L. Gradus, and A. J. Rosellini. 2020. “Supervised machine

learning: A brief primer.”Behav. Ther. 51 (5): 675–687. https://doi

.org/10.1016/j.beth.2020.05.002.

Kanter, R. M. 1981. “Power, leadership, and participatory management.”

Theor. Pract. 20 (4): 219–224. https://doi.org/10.1080/00405848109542958.

Kanter, R. M. 1993. Men and women of the corporation. 1993rd ed.

New York: Basic Books.

Kanter, R. M. 2021. “Look ‘outside the building’for the future of leader-

ship.”Leader Leader 2021 (100): 31–34. https://doi.org/10.1002/ltl

.20560.

Kettunen, J., and J. P. Makela. 2019. “Practitioners’conceptions of ethical

practice in social networking in career services.”Int. J. Educ. Voca-

tional Guidance 19 (3): 345–362. https://doi.org/10.1007/s10775-018

-9383-4.

Khalef, R., and I. H. El-adaway. 2021. “Automated identification of sub-

stantial changes in construction projects of airport improvement pro-

gram: Machine learning and natural language processing comparative

analysis.”J. Manage. Eng. 37 (6): 04021062. https://doi.org/10.1061

/(ASCE)ME.1943-5479.0000959.

Knowles, J., and L. Mainiero. 2021. “Authentic talent development in

women leaders who opted out: Discovering authenticity, balance, and

challenge through the kaleidoscope career model.”Administrative Sci.

11 (2): 60. https://doi.org/10.3390/admsci11020060.

J. Manage. Eng., 2022, 38(6): 04022060

Koch, M., B. Forgues, and V. Monties. 2017. “The way to the top: Career

patterns of Fortune 100 CEOs.”Hum. Resour. Manage. 56 (2): 267–285.

https://doi.org/10.1002/hrm.21759.

Lent, R. W., and S. D. Brown. 2019. “Social cognitive career theory at 25:

Empirical status of the interest, choice, and performance models.”

J. Vocational Behav. 115 (Dec): 103316. https://doi.org/10.1016/j

.jvb.2019.06.004.

Lent, R. W., S. D. Brown, and G. Hackett. 1994. “Toward a unifying social

cognitive theory of career and academic interest, choice, and perfor-

mance.”J. Vocational Behav. 45 (1): 79–122. https://doi.org/10.1006

/jvbe.1994.1027.

Li, L., A. Erfani, Y. Wang, and Q. Cui. 2021. “Anatomy into the battle

of supporting or opposing reopening amid the COVID-19 pandemic

on Twitter: A temporal and spatial analysis.”PLoS One 16 (7):

e0254359. https://doi.org/10.1371/journal.pone.0254359.

Liberty, E., K. Lang, and K. Shmakov. 2016. “Stratified sampling meets

machine learning.”In Proc., Int. Conf. on Machine Learning, 2320–

2329. Brookline, MA: Journal of Machine Learning Research.

Lin, N. 2001. Social capital: A theory of social structure and action (Ser.

Structural analysis in the social sciences, 19). Cambridge: Cambridge

University Press.

Lyness, K. S., and D. E. Thompson. 2000. “Climbing the corporate ladder:

Do female and male executives follow the same route?”J. Appl.

Psychol. 85 (1): 86–101. https://doi.org/10.1037/0021-9010.85.1.86.

Marinelli, M., S. Male, and L. Lord. 2021. “Women engineers’advance-

ment to management and leadership roles—Enabling resources and

implications for higher education.”In Proc., 2021 IEEE Global Engi-

neering Education Conf. (EDUCON), 463–467. New York: IEEE.

Maurer, J. A., D. Choi, and H. Hur. 2021. “Building a diverse engineering

and construction industry: Public and private sector retention of women

in the civil engineering workforce.”J. Manage. Eng. 37 (4): 04021028.

https://doi.org/10.1061/(ASCE)ME.1943-5479.0000913.

Miller, T., and M. Del Carmen Triana. 2009. “Demographic diversity in the

boardroom: Mediators of the board diversity-firm performance relation-

ship.”J. Manage. Stud. 46 (5): 755–786. https://doi.org/10.1111/j.1467

-6486.2009.00839.x.

Morgan, J. 2020. “Meeting the greatest challenges for leaders of the future.”

Leader Leader 2020 (96): 19–26. https://doi.org/10.1002/ltl.20500.

National Academy of Engineering. 2018. Understanding the educational

and career pathways of engineers. Washington, DC: National Acad-

emies Press.

Neumeyer, X., S. C. Santos, A. Caetano, and P. Kalbfleisch. 2019. “Entre-

preneurship ecosystems and women entrepreneurs: A social capital

and network approach.”Small Bus. Econ.: Entrepreneurship J. 53 (2):

475–489. https://doi.org/10.1007/s11187-018-9996-5.

Patel, J. M. 2020. “Web scraping in Python using BeautifulSoup library.”

In Getting structured data from the internet. Berkeley, CA: Apress.

Perrenoud, A. J., B. F. Bigelow, and E. M. Perkins. 2020. “Advancing

women in construction: Gender differences in attraction and retention

factors with managers in the electrical construction industry.”J. Man-

age. Eng. 36 (5): 04020043. https://doi.org/10.1061/(ASCE)ME.1943

-5479.0000808.

Poleacovschi, C., A. Javernick-Will, S. Wang, and T. Tong. 2021.

“Gendered knowledge accessibility: Evaluating the role of gender in

knowledge seeking among engineers in the US.”J. Manage. Eng.

37 (1): 04020094. https://doi.org/10.1061/(ASCE)ME.1943-5479

.0000865.

Powell, G. N. 2020. “Gender and leadership. ProQuest Ebook Central.”

Accessed July 15, 2021. https://ebookcentral.proquest.com.

Probst, P., M. N. Wright, and A. L. Boulesteix. 2019. “Hyperparameters

and tuning strategies for random forest.”Wiley Interdiscip. Rev.: Data

Min. Knowl. Discovery 9 (3): e1301. https://doi.org/10.1002/widm

.1301.

Ray, S. 2019. “A quick review of machine learning algorithms.”In Proc.,

2019 Int. Conf. on Machine Learning, Big Data, Cloud and Parallel

Computing (COMITCon),35–39. New York: IEEE.

Roebuck, D. B., and D. N. Smith. 2013. “Wisdom from executive female

leaders: What can organizations, executive education programs, and

graduate students learn?”J. Executive Educ. 10 (1): 3.

Sandberg, S. 2013. Lean in: Women, work, and the will to lead. New York:

Knopf.

Schultheiss, D. E. 2021. “Shining the light on women’s work, this time

brighter: Let’s start at the top.”J. Vocational Behav. 126 (Apr):

103558. https://doi.org/10.1016/j.jvb.2021.103558.

Seger, C. 2018. An investigation of categorical variable encoding tech-

niques in machine learning: Binary versus one-hot and feature hashing.

Stockholm, Sweden: KTH, School of Electrical Engineering and

Computer Science (EECS).

Seibert, S. E., M. L. Kraimer, and R. C. Liden. 2001. “A social capital

theory of career success.”Acad. Manage. J. 44 (2): 219–237. https://doi

.org/10.2307/3069452.

Seyrfar, A., H. Ataei, A. Movahedi, and S. Derrible. 2021. “Data-driven

approach for evaluating the energy efficiency in multifamily residential

buildings.”Pract. Period. Struct. Des. Constr. 26 (2): 04020074.

https://doi.org/10.1061/(ASCE)SC.1943-5576.0000555.

Sunindijo, R., and I. Kamardeen. 2017. “Work stress is a threat to gender

diversity in the constructionindustry.”J. Constr. Eng. Manage. 143 (10):

04017073. https://doi.org/10.1061/(ASCE)CO.1943-7862.0001387.

Thomas, R. R., Jr. 1990. “From affirmative action to affirming diversity.”

Harvard Bus. Rev. 68 (2): 107–117.

True-Funk, A., C. Poleacovschi, G. Jones-Johnson, S. Feinstein, K. Smith,

and S. Luster-Teasley. 2021. “Intersectional engineers: Diversity of

gender and race microaggressions and their effects in engineering

education.”J. Manage. Eng. 37 (3): 04021002. https://doi.org/10.1061

/(ASCE)ME.1943-5479.0000889.

US Bureau of Labor Statistics. 2020. “Labor force statistics from the cur-

rent population survey.”Accessed July 7, 2021. https://www.bls.gov

/cps/cpsaat18.htm.

US Equal Employment Opportunity Commission. 2021. “Title VII of the

Civil Rights Act of 1964.”Accessed August 10, 2021. https://www.eeoc

.gov/statutes/title-vii-civil-rights-act-1964.

Yoder, B. 2019. “Engineering by the numbers.”In American society of

engineering education, reports from 2000–2018. Washington, DC:

American Society for Engineering Education.

Yousaf, S., A. A. Humayon, M. I. Rasheed, H. M. Ahmed, and R. Q.

Danish. 2014. “Factors affecting retention of female employees in

organizations.”J. Basic Appl. Sci. Res. 4 (7): 98–106.

Zhang, C., and Y. Ma. 2012. Ensemble machine learning: Methods and

applications. Boston, MA: Springer Science and Business Media.

Zhang, J. Q., H. Zhu, and H.-B. Ding. 2013. “Board composition and

corporate social responsibility: An empirical investigation in the post

Sarbanes Oxley era.”J. Bus. Ethics 114 (3): 381–392. https://doi.org/10

.1007/s10551-012-1352-0.

J. Manage. Eng., 2022, 38(6): 04022060

State-of-the-Art Review Paving the Way for Progress: A Systematic Literature Review on Diversity, Equity, and Inclusion in the AEC Industry

Article

Full-text available

Apr 2024
J MANAGE ENG

The importance of diversity, equity and inclusion (DEI) in the architecture, engineering, and construction (AEC) industry has become increasingly salient as the industry continues to struggle with recruitment and retention of high-quality talent throughout construction organizations in the US. The existing literature has explored the barriers faced by various groups who are underrepresented in the construction industry; however, the research is fragmented and lacks a set of synthetic conclusions to drive future work. To this end, we performed a systematic literature review (SLR) to analyze the existing literature, which has categorized the barriers contributing to underrepresentation of women, ethnic minorities, individuals with disabilities, individuals who identify as lesbian, gay, bisexual, transgender, and queer/questioning (LGBTQ), and individuals who have been previously incarcerated. As a result of the SLR, 10 influential variable clusters affecting the recruitment and retention of underrepresented groups were identified, including (1) organizational aspects, (2) work environment, (3) work-life balance, (4) individual or project performance, (5) education, (6) health, (7) safety, (8) wage gap, (9) communication, and (10) ethics. Our findings lead to suggestions for further research, including a need to identify best practices based on empirical evidence from within the range of current DEI practices that have been investigated by prior research. Our results can benefit practitioners, legislators, employers, and others working to improve DEI in the construction industry.

Crowdfunding for Equitable EV Charging Infrastructure

Preprint

Full-text available

Jun 2024

The transportation sector significantly contributes to greenhouse gas emissions, highlighting the need to transition to Electric Vehicles (EVs) to reduce fossil fuel dependence and combat climate change. The US government has set ambitious targets for 2030, aiming for half of all new vehicles sold to be zero-emissions. Expanding EV charging stations is crucial for this transition, but social equity presents a significant challenge. The Justice40 program mandates that at least 40% of benefits be allocated to disadvantaged communities, ensuring they benefit from federal investments. Given the current concentration of EV ownership in affluent areas, merely installing charging stations in disadvantaged neighborhoods may not suffice. This article explores crowdfunding as a novel method to finance EV charging infrastructure, engaging, and empowering underserved communities. The paper concludes with a hypothetical case showing financing benefits for disadvantaged communities, exploring crowdfunding variations, and scaling to develop equitable EV charging networks.

Analysis of Gender Diversity Initiatives to Empower Women in the Australian Construction Industry

Article

Full-text available

Jun 2024

Organizations play an essential role in fostering a diverse and inclusive work environment. Despite attempts to encourage gender diversity in the workplace, the construction industry in Australia remains one of the most male-dominated industries. Existing research tends to focus on identifying barriers to gender equality and women’s career development, with limited exploration of organizations’ current initiative implementation. To fill this research gap, this study aims to examine how organizations implement initiatives affecting gender equality and women’s career development in the Australian construction industry. This study applied the women’s empowerment framework and conducted a thematic analysis of diversity, inclusion, and equality policies, strategies, action plans, corporate reports, annual reports, and websites within the context of gender equality and women’s career development in construction companies. Through a comprehensive analysis, the study identifies prevalent practices and initiatives related to gender equality initiatives in the construction industry, synthesizing them into six themes: flexible working arrangements, inclusive and anti-discriminatory culture, promoting career development, recruitment, and promotion, shared caring responsibility, and strategic commitments. The findings also revealed areas for improvement, giving recommendations on how to further empower women in the construction industry through enhanced gender diversity initiatives. This study sheds light on the current state of gender diversity and inclusion efforts within construction companies and provides practical recommendations for policymakers, industry stakeholders, and organizational leaders who seek to foster more inclusive and equitable work environments.

Career Interest Assessment: College Students Career Planning Based On Machine Leaning

Article

Full-text available

Apr 2024

Can Cui

College students' career planning based on career interest assessment and machine learning integrates advanced technological tools with traditional career counseling methods to provide personalized and data-driven guidance. By utilizing career interest assessment tools, students can explore their strengths, preferences, and aspirations across various career domains. Machine learning algorithms analyze this data to generate tailored recommendations, matching students with career paths that align with their interests and aptitudes. This paper presents a novel approach to career assessment and planning by leveraging the Sugeno Fuzzy Optimized Weighted Machine Learning (SFOwML) framework. The proposed framework integrates fuzzy logic principles with machine learning techniques to provide personalized and accurate career recommendations tailored to individuals' skills, preferences, and values. Through the incorporation of optimization techniques, the framework refines the career assessment model, enhancing its accuracy in predicting suitable career paths for individuals. The practical application of the SFOwML framework offers a valuable tool for career counselors, educators, and individuals seeking guidance in their career choices. The framework incorporates numerical values to quantify individuals' skills, preferences, and values on a scale from 0 to 100, allowing for a more precise evaluation of their career profiles. Through the integration of fuzzy logic principles with machine learning techniques, personalized career recommendations are generated based on these numerical assessments.

Stacked Ranking Feature Cluster Machine Learning (Srfcml): A Novel Method of Career Planning of College Students Based on Career Interest Assessment and Machine Learning

Article

Full-text available

Apr 2024

Jing Lv

Career interest assessment, powered by machine learning algorithms, revolutionizes the way individuals explore and align with career paths. By analyzing vast datasets encompassing factors such as skills, preferences, personality traits, and job market trends, machine learning models can provide personalized career recommendations tailored to individual strengths and aspirations. These algorithms leverage advanced techniques such as natural language processing (NLP) to interpret self-assessment responses and match them with suitable career options. Additionally, machine learning algorithms continuously refine their recommendations based on user feedback and real-world outcomes, ensuring accuracy and relevance over time. This paper presents a novel approach to career planning for college students, integrating career interest assessment with machine learning techniques, specifically utilizing the Stacked Ranking Feature Cluster Machine Learning (SRFcML) model. The proposed framework leverages large-scale datasets encompassing diverse factors such as academic performance, skills, interests, and industry trends to provide personalized career recommendations. Through the application of machine learning algorithms, including clustering and ranking techniques, the SRFcML model identifies relevant clusters of career options and ranks them based on individual preferences and aptitudes. This approach enables college students to explore and prioritize career paths aligned with their unique strengths and aspirations. Simualtion results demonstrated that the effectiveness of the proposed framework was evaluated, yielding promising numerical results. For instance, based on self-assessment responses and academic performance data, the SRFcML model achieved an average accuracy of 85% in recommending suitable career paths for college students. Furthermore, in a comparison with traditional career planning methods, the proposed framework demonstrated a 30% improvement in the alignment between recommended career options and students' preferences. Additionally, user satisfaction surveys revealed a high level of confidence and trust in the recommendations provided by the SRFcML model, with 90% of participants expressing satisfaction with the accuracy and relevance of the career suggestions.

Measuring Social Equity in Pavement Conditions Using Big Data

Conference Paper

Full-text available

Mar 2024

Unconscious Bias in Self-Presentation: Exploring Gender Disparities in Leadership within the Transportation Industry

Conference Paper

Jun 2024

Understanding Current Demand for BIM Professionals in China through Recruitment Data Mining

Article

May 2024

Cracking the Diversity Code: Analyzing Board Gender Diversity in the AEC Sector

Article

May 2024

Eunpyo Hong

Listening to Leaders—How AEC Women Grow Their Career Pathways

Conference Paper

Mar 2024

Occupational commitment of women working in SET: The impact of coping self‐efficacy and mentoring

Article

Full-text available

Aug 2021
Hum Resource Manag J

The under-representation of women employed in Science, Engineering and Technology (SET) industries is a long-term problem for human resource management. We report the results of a qualitative research study designed to investigate the factors that positively influence women's retention and occupational commitment within the MENA region. The data are based on 40 semi-structured interviews of women employed in SET occupations. The results reveal that self-directed attitudes (coping self-efficacy, protean attitude and professional identity) and contextual support (mentoring and the quality of the mentoring relationship) positively influence women's occupational commitment in jobs in SET. Our contributions to academic research on HRM and HRD include the important role that the psychological functions of mentoring play in OCC. In terms of the contribution to practice, we propose that professional HRM institutes and associations could contribute by launching campaigns to promote greater organisational awareness of the potential of mentoring for retaining women in SET.

Women’s journey to the top: the role of bonding and bridging social capital

Article

Full-text available

Aug 2021

Purpose Drawing on Putnam’s concept of social capital, this study aims to examine the oral life histories of women on management boards. It explores how the bridging and bonding forms of social capital are created and used to advance women’s careers to reach top management positions on corporate boards. Design/methodology/approach This study is both explorative and interpretive. A total of 30 semi-structured interviews were conducted with female board members of Polish companies to gain access to their life histories. Findings Women develop social capital to advance their careers. However, they mainly focus on bonding capital, which allows them to sustain relationships within their inner circle. Bridging capital, which extends relationships to advance women’s careers, is often overlooked if not neglected. It seems it is usually an external shock when women understand the importance of bridging capital. Practical implications This study provides a better insight into how bridging and bonding capital help and/or distract women from reaching top managerial positions. It has potential policy-making implications for promoting women to leadership positions. The study results can guide organizations in providing equal opportunities for employees. Originality/value This study builds on previous research regarding the accumulation of human and social capital by board members. It encourages academics and practitioners to critically consider the relationship between bonding and bridging capital and the promotion of women to top management positions. This study provides access to a field of Central and Eastern European countries.

Anatomy into the battle of supporting or opposing reopening amid the COVID-19 pandemic on Twitter: A temporal and spatial analysis

Article

Full-text available

Jul 2021
PLOS ONE

Reopening amid the COVID-19 pandemic has triggered a battle on social media. The supporters perceived that the lockdown policy could damage the economy and exacerbate social inequality. By contrast, the opponents believed it was necessary to contain the spread and ensure a safe environment for recovery. Anatomy into the battle is of importance to address public concerns, beliefs, and values, thereby enabling policymakers to determine the appropriate solutions to implement reopening policy. To this end, we investigated over 1.5 million related Twitter postings from April 17 to May 30, 2020. With the aid of natural language processing (NLP) techniques and machine learning classifiers, we classified each tweet into either a “supporting” or “opposing” class and then investigated the public perception from temporal and spatial perspectives. From the temporal dimension, we found that both political and scientific news that were extensively discussed on Twitter led to the perception of opposing reopening. Further, being the first mover with full reopen adversely affected the public reaction to reopening policy, while being the follower or late mover resulted in positive responses. From the spatial dimension, the correlation and regression analyses suggest that the state-level perception was very likely to be associated with political affiliation and health value.

Authentic Talent Development in Women Leaders Who Opted Out: Discovering Authenticity, Balance, and Challenge through the Kaleidoscope Career Model

Article

Full-text available

Jun 2021

We investigated the parameters of authenticity, balance, and challenge as viewed through the lens of the Kaleidoscope Career Model to discern the career motives of women who opted out and then returned to the workforce. We also were interested in comparing women and men in their leader practices. We triangulated quantitative and qualitative methods to explore these phenomena. First, in Study 1, 2009 individuals completed both the Kaleidoscope Career Self Inventory (the KCSI) as well as an abbreviated version of Kouzes and Posner’s Leadership Practices Inventory (the LPI). Participants rated their needs for authenticity, balance, and challenge on the KCSI as well as their leader practices of challenging the process, inspiring a shared vision, modeling the way, enabling others to act, and Encouraging the Heart. Women were found to exceed the leader practices of men, and women were most interested in authenticity. In Study 2, situational mapping and life history process was used to determine themes of eight high-achieving women who opted out. We interviewed an additional 15 women to deeply understand and assess their opt-out and re-entry experiences as well as “career shocks” they experienced upon returning to the workforce. Our second study offers a robust, deep, penetrating look into social ascription processes and endemic discriminatory social structures that hold women back from achieving advancement. To stop this “brain drain” of talent, we propose a series of actions for human resource professionals to develop the authentic leadership talent of women who reenter the workforce.

Gendered Knowledge Accessibility: Evaluating the Role of Gender in Knowledge Seeking among Engineers in the US

Article

Jan 2021

Automated Identification of Substantial Changes in Construction Projects of Airport Improvement Program: Machine Learning and Natural Language Processing Comparative Analysis

Article

Nov 2021

Contractual changes - mainly substantial changes - within airport improvement program (AIP) projects represent a critical risk that could result in severe negative time and cost impacts. It is critical for airport projects to have in place efficient procedures to process changes effectively, or otherwise this may create an administrative choke point for their stakeholders. Further, with the current US airport infrastructure scoring a D+ (i.e., lacking behind the general US infrastructure), associated authorities called for rebuilding the US airport infrastructure. Thus, it is expected that contractual changes are going to increase for current as well as future US airport projects. This makes it critical to identify these changes early on to incorporate proper change management strategies. However, analysis of contract documents is a process that is known to be inefficient, tedious, and prone to human error. The goal of this research is to create an automated framework to predict substantial contractual changes effectively and efficiently within AIP construction projects. An independent multistep research methodology was used based on principles of natural language processing (NLP) and machine learning techniques (ML). First, the authors adopted a data set containing 876 contractual changes made to the Federal Aviation Administration (FAA) document of guidelines and policies that govern AIP projects (FAA 5100.38D). Second, the authors used NLP techniques to preprocess the aforementioned data. Third, the authors developed hyperparameter-tuned ML models, including k-nearest neighbor (KNN), support vector machine (SVM), random forest (RF), artificial neural network (ANN), extreme gradient boosting (XGBoost), and logistic regression (LR) to predict substantial changes made to the FAA 5100.38D. Accordingly, results indicate that RF showed the most accurate prediction with an area under curve (AUC) value of 0.928, a testing accuracy of 87.45%, and a mean cross-validation accuracy of 92.67%. As such, this automated framework grants stakeholders associated with AIP construction projects a computational decision support tool to easily recognize substantial changes within contract documents, both efficiently and effectively. Ultimately, this research promotes better change management implementation and supports overall AIP project success.

Building a Diverse Engineering and Construction Industry: Public and Private Sector Retention of Women in the Civil Engineering Workforce

Article

Jul 2021

The gender gap in the construction and engineering workforce persists, despite efforts to close it over the past several decades. Increasing gender diversity is a critical strategy for strengthening the supply of engineering professionals required to meet current and future demand in both public and private sectors. This study evaluates major demographic trends in the civil engineering workforce during the past decade with an emphasis on gender diversity. We also examine trends affecting the retention of female supervisors by examining job turnover in both public and private sectors. We found that female civil engineers are underrepresented in supervisory positions as compared to their representation in the workforce overall. They are also much more likely to leave their jobs than male supervisors. Their departure rate is higher in the private sector than in the public sector. These results illustrate the advancement challenges that women encounter on their career paths, and the consequences of these barriers - namely, less gender diversity in the pool of civil engineers available to meet growing demand in the construction industry, and fewer women available for promotion into leadership positions. The findings of this study will be beneficial to managers and other stakeholders working to diversify the construction industry workforce.

LOOK “OUTSIDE THE BUILDING” FOR THE FUTURE OF LEADERSHIP

Article

Mar 2021

Rosabeth Moss Kanter

The author discusses what she terms “outside the building” issues, such as climate change, pandemics, and racial/gender justice. These complex issues are “beyond the boundaries of single organizations.” As she writes, “I call this new leadership for the future ’advanced’ leadership, because of its broader scope and, at the same time, more limited formal power, which challenges leaders more than an official mandate does.” She references former First Lady of the United States Michelle Obama, the late leadership author Warren Bennis, and the academics Robert Putnam (famous for his book Bowling Alone), Sherry Turkle, and others. The author observes that “Innovation and change can’t follow a fixed script. The work is more like improvisational theater, which takes shape in response to observer feedback.” While it may seem that society is hopelessly divided and gridlocked, there are hopeful signs: “When more people flex their leadership muscles and find the courage to take on big problems, they exude the optimism of activism. That positive energy can be contagious and can unite us in new ways. By undertaking the daunting, yet meaningful, task of making a difference in the world, advanced leaders can transform the lives of many people—including the leaders themselves.”

Sex, Age, Work Experience, and Relatives in Building Engineering Career Development

Article

May 2021

Leadership Diversity and its Influence on Equality Initiatives and Performance: Insights for Construction Management (A*; In press)

Article

May 2021

Gender-diverse leadership shapes the design and delivery of policies, programs, and practices implemented to address diversity management. Drawing on stakeholder theory, this paper proposes and tests the positive impacts of board gender diversity and top management team gender diversity on diversity-related human resource initiatives and work–life initiatives. Moreover, using strategic human resource management theory and ability–motivation–opportunity theory, this study proposes and tests the positive effects of human resource and work–life initiatives on organizational financial performance. Based on contingency theory, it proposes and tests the moderating effect of an inclusive diversity environment on the initiatives and financial performance relationship. Using time-lagged archival data collected from 367 Australian organizations, findings indicate board gender diversity is positively associated with both human resource and work–life initiatives designed and implemented to address diversity management. Work–life initiatives are positively associated with both increased operating revenue and profit before tax. Further, the work–life initiatives–profit before tax relationship was strongly positive for organizations with a high diversity environment. We discuss theoretical and research contributions and practical implications.

Use of LinkedIn Data and Machine Learning to Analyze Gender Differences in Construction Career Paths

Abstract and Figures

Recommended publications

A Gender Analysis of Construction Executive Career Path Progression Using LinkedIn Data

Likeability versus Competence Dilemma: Text Mining Approach Using LinkedIn Data

Listening to Leaders—How AEC Women Grow Their Career Pathways

Application of LinkedIn Data and Image Processing to Analyze Construction Career Paths: Does Race Ma...