Content uploaded by Klaas-Jan Stol
Author content
All content in this area was uploaded by Klaas-Jan Stol on Oct 25, 2020
Content may be subject to copyright.
The Journal of Systems and Software 159 (2020) 110442
Contents lists available at ScienceDirect
The Journal of Systems and Software
journal homepage: www.elsevier.com/locate/jss
Exploring onboarding success, organizational fit, and turnover
intention of software professionals
Gaurav G. Sharmaa,c, Klaas-Jan Stolb,c,∗
aIBM, Cork, Ireland
bLero—the Irish Software Research Centre, Ireland
cSchool of Computer Science and Information Technology, University College Cork, Ireland
ARTICLE INFO
Article history:
Received 17 March 2019
Revised 9 October 2019
Accepted 10 October 2019
Available online 11 October 2019
Keywords:
Onboarding
job satisfaction
turnover intention
survey
PLS
ABSTRACT
The IT sector struggles with talent acquisition and low retention rates. While several field studies have ex-
plored onboarding of software developers, the software engineering literature lacks studies that develop
and evaluate theoretical models. This study seeks to explore the link between onboarding of new hires and
turnover intention of these professionals. In particular, we develop a theoretical model that identifies a num-
ber of onboarding activities, and link these to onboarding success. We then look at what we have termed
organizational fit, which we define as two aspects of software professionals, namely job satisfaction and
the quality of their relationships on the workfloor, and investigate how these mediate the relation between
short-term onboarding success and a longer-term intention to leave (or stay with) an organization. We test
our model with a sample of 102 software professionals using PLS-SEM. The findings suggest that providing
support to new hires plays a majorrole in onboarding success, but that training is less important. Further, we
found that job satisfaction mediates the relationship between onboarding success and turnover intention,
but workplace relationship quality does not. Based on the findings, we discuss a number of implications for
practice and suggestions for future research.
1. Introduction
Software professionals are extremely mobile today’s highly in-
terconnected world (Forrest, 2018). Developers could potentially
work from any place with an Internet connection, and can there-
fore change jobs very easily. Besides this type of flexibility, it is
quite common for developers to move to a new company every
few years or so (Miller, 2018). Each time an organization hires a
new software developer, this person must be introduced to the
organization, its processes, and its culture. Typically this is done
through an onboarding process. Recruiting and onboarding people
with the right skills and personality is crucial to the success of soft-
ware development organizations and projects (Hall et al., 2008).
Recruitment, however, is an expensive activity; it is not unusual
that newly recruited developers are initially a liability since there
is a learning curve that any employee goes through before becom-
ing productive (Brooks, 1975;DeMarco and Lister, 1987). Some ex-
perts suggest it may take up to 6 to 12 months before new recruits
become productive (Sim and Holt, 1998).
∗Corresponding author at: School of Computer Science and Information
Technology, University College Cork, Ireland.
E-mail address: klaas-jan.stol@lero.ie (K. Stol)
In addition to the productivity delay, there is also a chasm
between the skills and knowledge that new graduates formal ed-
ucation offers, and what industry requires (Brechner, 2003;Legier
et al., 2013;Radermacher and Walia, 2013;Tang et al., 2001;Trent,
1988). Technology is changing constantly (Trent, 1988), and while
the technical challenges that newcomers face can be significant,
non-technical skills have also consistently been found to be very
important and highly rated skills by employers (Aasheim et al.,
2009;Begel and Simon, 2008a;Legier et al., 2013;McMurtrey et
al., 2008;Simon and Jackson, 2013;Tesch et al., 2008).
Onboarding, also known as organizational socialization in
the management literature (Bauer, 2010;Van Maanen and Schein,
1979), is a formal or informal process of integrating newly hired
employees and transforming them from being outsidersto pro-
ductive members of the organization. This involves the transfer
of knowledge, skills, rules, and familiarity with the organizational
culture to be able to work within a team (Brittoet al., 2018;Sim and
Holt, 1998). Research suggests that the first 90 days are crucial and
decide the success of a newcomer in his or her job (Watkins, 2013),
and it is typical that onboarding activities take place during this
initial period. This period is also important from a financial per-
spective, because hiring and onboarding new employees is a costly
process. The costs associated with onboarding new employees are
https://doi.org/10.1016/j.jss.2019.110442
This PDF was created with L
A
T
EX and is not the publisher’s version.
2G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
due to an initial low level of productivity, and HR administration
and bureaucracy involved in hiring new staff (Snell, 2006). In the
context of software development, developing an understanding of
an unfamiliar codebase written by others can be a complicated task
for most software developers (LaToza et al., 2006). Therefore, the
importance of the onboarding process looms even larger for multi-
national software companies working across the globe and which
often depend on legacy systems. A lack of an effective plan could
cause problems in maintaining the software (Littman et al., 1987),
which could lead to a longer lead time to full productivity (Sim and
Holt, 1998).
It is worth clarifying what we mean by the term onboarding.
In this article we adopt the term to mean the same as in the man-
agement literature cited above, namely the process of organiza-
tional socialization. This is consistent with Microsofts use of the
term defined by Begel and Simon (2008a), who define onboar-
ding as the Microsoft term for the orientation process by which
new hires adjust to and become effective software developers within
the corporation. Other metaphors to characterize newcomers and
their journey have been used, such as explorers who must orient
themselves within an unfamiliar landscape (Dagenais et al., 2010),
ramp-up journey (Rastogi et al., 2015), and software immigrants
who must adapt through a process of naturalization (Sim and
Holt, 1998). Following Sim and Holt (1998), we note that the term
novice may be inappropriate because it implies that a new re-
cruit is a new graduate, but clearly new recruits may have exten-
sive industry experience when they come from other organiza-
tions. Hence, we adopt the term newcomer in this article.
The interpretation of the term onboarding described above is
a holistic process of settling in, rather than the more narrow mean-
ing of becoming familiar with a specific software code base which
has also been used by some software engineering researchers (cf.
Yates et al. (2020)). There is a considerable body of literature that
focuses on understanding legacy systems and program compre-
hension, pioneered primarily by the software maintenance com-
munity (represented by the International Conference on Software
Maintenance and Evolution (ICSME) and related journals such as
Journal of Software: Evolution and Process) (Von Mayrhauser et al.,
1997). Specific techniques include information seeking (O’Brien,
2007), feature location (Dit et al., 2013), source code analysis (Sil-
lito et al., 2008), reading software documentation (Lethbridge et al.,
2003), and querying knowledgeable peers with a longer tenure at
the organization (Hertzum and Pejtersen, 2000). While these tech-
niques are related to onboarding software developers in a more
specific sense, in this article we focus on the more holistic process
of onboarding suggested above.
There are many studies in the management literature that
discuss strategies and best practices in general—including Van
Maanen and Schein (1979)s seminal Theory of Organizational
Socialization—but few focus specifically on software professionals.
The software engineering domain faces particular challenges given
regular reports of shortages of software engineers, briefly men-
tioned above, including newcomers productivity delay, a fast rate
of technology change, and a high level of turnover of developers.
Some media have reported that the turnover in the software sec-
tor is the highest of all (Forrest, 2018). A recent study by Gupta
et al. (2018) investigating the relationship between new hires on-
boarding experience and their intention to leave (what we refer to
as turnover intention) found that the latter was highest for the
IT sector. Hence, this suggests the need for further studies that
explore this relationship in the IT sector. Another reason to fo-
cus specifically on the software sector is that new hires, or what
Sim and Holt (1998) have labeled software immigrants, must
acquire a wide variety of knowledge in order to become produc-
tive. Besides general knowledge such as programming languages
and tools, company-specific knowledge must be acquired such as
project jargon, team dynamics, coding standards, and organiza-
tional structures (Sim and Holt, 1998;Hilton and Begel, 2018).
Within the software engineering literature, most studies of on-
boarding focus on attracting and sustaining of new contributors in
open source software communities (e.g. (Steinmacher et al., 2014;
Gharehyazie et al., 2015)), and relatively limited attention to the
onboarding process of software professionals. There are a few field
studies on onboarding practices and techniques adopted by glob-
ally distributed companies such as Google (Johnson and Senges,
2010) and Microsoft (Begel and Simon, 2008b;Begel and Simon,
2008a) (see Table 1 for an overview). While field studies provide
rich contextual insights, the findings of such studies are inherently
limited in generalizability (Stol and Fitzgerald, 2018)—thesoftware
engineering literature lacks tested theories that help explain what
makes the onboarding process successful, and its potential influ-
ence on professionals sense of what we term organizational fit,
and ultimately their intention to stay with or leave their organiza-
tion. In this article, we draw on the wider literature from several re-
search fields, including organizational, management and psychol-
ogy literature, to synthesize a theoretical model to investigate this.
The goal of this study is to develop insights regarding which factors
might help to achieve a successful onboarding experience, how the
onboarding experience relates to developers settling in in the or-
ganization, and how a lack of fit might increase peoples inten-
tion to leave the organization.
The remainder of this article is organized as follows. In Sec-
tion 2 we develop a theoretical model that informs our empirical
study to test our theory—Section 3 presents details on our research
strategy, followed by the study results in Section 4.Section 5 dis-
cusses our findings, limitations of the study, and concludes this
article.
2. Theory development
In this section we review prior work on onboarding in com-
mercial software development organizations, and draw on man-
agement and psychology literature to develop a theoretical model.
Table 1 providesan overview of previous studies of onboarding in a
commercial software engineering context. As we focus specifically
on onboarding in companies, this overview does not include the
growing body of literature on onboarding in open source projects
(cf. Steinmacher et al., 2014;Fagerholm et al., 2014b;Casalnuovo
et al., 2015). Neither does the table include studies that focus ex-
clusively on specific activities such as knowledge transfer between
senior and novice developers (cf. Viana et al., 2014)—while related,
the focus of our study is specifically on onboarding.
2.1. Onboarding activities
Onboarding of new employees involves a variety of activities.
Van Maanen and Schein (1979), who refer to onboarding as organi-
zational socialization described this as:
the process by which one is taught and learns the ropes of a
particular organizational role. In its most general sense, orga-
nizational socialization is then the process by which an indi-
vidual acquires the social knowledge and skills necessary to as-
sume an organizational role.
In terms of the social knowledge and skills that Van Maanen
and Schein (1979) refer to, we focus specifically on three types of
activities which are recurrent themes in prior studies (see Table 1):
orientation, involving introducing a newcomer to the organization
(Begel and Simon, 2008b;Begel and Simon, 2008a;Britto et al.,
2018); training, which focuses on providing sufficient information
to a newcomer to do their job (Begel and Simon, 2008b;Begel and
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 3
Table 1
Selection of prior empirical work on onboarding in commercial software development organizations
Study Method Findings
Berlin (1993) Field study at HP Labs comprising observations of
three expert-apprentice pairs and interviews to
explore the complex ways in which experts help
apprentices.
Learning new development environments and tools can be a major barrier to
productivity. Experts tend to ask others for help more readily than
apprentices, as the latter may not want to overburden their mentor.
Sim and Holt (1998) Longitudinal field study that seeks to understand
the naturalization process of software
immigrants, based on periodic interviews with
4 developers working on a compiler component
system, at a very large computer company.
Set of seven observed patterns of the naturalization process, organized in four
categories, relating to (1) mentoring, (2) difficulties unrelated to the software
system being learned, (3) first assignments, and (4) job fit. An example
pattern observed is: mentors are an effective, though inefficient way to teach
immigrants about the software system.
Begel and Simon
(2008b) and Begel
and Simon (2008a)
Field study comprising 85 hours of observation of
eight recent college graduates during their first
six months at Microsoft Corp., studying
newcomers activities, interactions, and
challenges.
Subjects strongest skills include the ability to write code, design specifications,
and persistence when facing difficult (technical) problems. Subjects struggled
with knowing when and how to ask questions, team interaction skills,
technical difficulties related to tools and testing, organizing and managing a
wide range of information, and orientation issues in the project (tools, code,
people).
Johnson and Senges
(2010)
Field study employing case study methodology at
Google, involving interviews with 24
stakeholders, observations, documents,
addressing the question: how is practice-based
learning incorporated in the onboarding process
of new software engineers?
Nooglers (new Google employees) receive two week face-to-face training and
orientation on core technologies and practices; senior engineers share
engineering values and language; online training including checklists,
tutorials and codewalks; on-the-job training includes a starter project.
Performance feedback is given on activities, objectives, and also through code
reviews.
Dagenais et al. (2010) Qualitative survey analyzed with a Grounded
Theory approach involving 18 newcomers from
18 projects at IBM, investigating: what are the
key, prominent features in a project landscape,
what orientation obstacles do new team
members face, and what orientation aids can be
provided?
Newcomers must learn the project landscape, with interactions and challenges
in five categories: product (incl. design, technologies); processes and
practices (incl. tools), team (incl. roles and expertise), documentation, and
context (incl. inter-team organization). Newcomers settle into a project
landscape through early experimentation, internalization of structures and
cultures, and progress validation (feedback).
Rastogi et al. (2015) Field study of 8 large projects at Microsoft, using
quantitative data from a version control system
and qualitative data from 4 interviews with
developers to investigate the ramp-up journey of
newly hired developers.
Lack of documentation, getting set up (i.e. access and permission, system
set-up), and lack of technical skills inhibit productivity. Mentorship was
highlighted to be important by managers.
Pham et al. (2017) Sample study using three online questionnaires
(with 54, 170, and 698 respondents); 22
interviews with developers, to investigate
practitioners perceptions of novice developers.
Software practitioners perceive a skill gap between university graduates and
industry expectations in relation to testing skills. Training and mentoring
efforts are expended to address this gap.
Britto et al. (2018) Field study of three globally distributed legacy
projects involving teams based in India, Norway,
Poland, Sweden, and the USA, investigating
onboarding and associated challenges.
Onboarding strategies vary across companies and even among different sites
within the same company. Onboarding newcomers onto projects with legacy
code is challenging when original developers are not onsite. Orientation was
neglected in all three case projects. Coaching and mentoring are most
prevalent practices, but this reduces mentors’ productivity.
Yates et al. (2020) Field study using a Grounded Theory methodology
of onboarding sessions across eight different
organizations.
Experts describe the code from four different viewpoints: a structural, an
algorithmic, a rationale, and a temporal view. Onboarding sessions facilitate
the transfer of knowledge that cannot be found in the code or documentation.
Simon, 2008a;Berlin, 1993;Johnson and Senges, 2010;Pham et
al., 2017;Rastogi et al., 2015); and support, which involves a set of
mechanisms to help, guide, and provide feedback to a newcomer
(Dagenais et al., 2010;Rastogi et al., 2015;Sim and Holt, 1998).
Van Maanen (1978) characterized organizational socialization
along six dimensions, two of which are relevant here. The first
is collective vs. individual socialization. In a collective process, a
group of newcomers are subjected to a common set of experiences
together (Van Maanen and Schein, 1979), whereas in an individ-
ual process, each newcomer has a unique experience. The second
dimensions is formal vs. informal. A formal socialization process
is one in which newcomers are separated from other employees
as they are subjected to a program specifically tailored to them.
An informal process does not differentiate newcomers, and in such
cases newcomers are expected to learn on the job. The remain-
ing four dimensions describe other aspects, such as whether the
process is sequential vs. random and fixed vs. variable. As we did
not draw on these dimensions in this study, we do not discuss these
here.
2.1.1. Orientation
Joining a new workplace comes with its share of stress and anx-
iety for a newcomer (Bourne, 1967;VanMaanen and Schein, 1979);
this initial stress and anxiety may inhibit software professionals
from becoming productive. Orientation programs should include
emotion-focused methods, along with problem-focused methods,
to reduce stress. Klein and Weaver (2000) defined orientation pro-
grams for newcomers to introduce them to their job, co-workers,
and the larger organization. Most orientation programs cover the
following three areas: (1) terms and conditions of employment,
(2) health, safety and legal issues, and (3) the organizations his-
tory, culture, and values (Wanous and Reichers, 2000). Orienta-
tion, also termed early socialization, typically takes place within
the first month of an employee joining an organization (Wanous
and Reichers, 2000;Anderson et al., 1996).
There are three accepted frameworks which guide the research
and design of orientation programs: (1) stress theory / coping
methods, (2) attitude theory / change, and (3) Realistic Job Preview
(RJP) theory. Different industries use one or a more of these to de-
velop orientation programs for their newcomers (Wanous and Re-
ichers, 2000).
4G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
Early studies of orientation programs showed few significant
correlations with long-term success factors, such as job satisfac-
tion and organizational commitment, self-efficacy, and coping
ability (Anderson et al., 1996;Bolles, 2000;Louis et al., 1983;
Nelson and Quick, 1991;Saks, 1995). However, these programs
remain one of the most popular methods for early socialization
across organizations (Feldman, 1989) because they help convey
compliance requirements and promote a positive image of the or-
ganization (Fan et al., 2012). Several studies have demonstrated
the benefits of attending formal orientation programs (Gundry,
1993;Klein and Weaver, 2000;Wesson and Gogus, 2005).
Building on this previous research, we argue there is a positive
link between the organization of orientation activities and what
Bauer (2010) has defined as onboarding success, namely knowl-
edge of the organizational culture, role clarity, self-efficacy, and
social integration. Further, orientation programs have also been
linked to a reduced level of stress in newcomers. We hypothesize
that:
Hypothesis 1 (H1). Orientation programs for newly recruited soft-
ware professionals have a positive association with onboarding suc-
cess.
2.1.2. Training
Whereas orientation is concerned with context performance
(focusing on the organization and its culture), training focuses on
task performance (focused on the tasks that the new recruit is ex-
pected to perform) (Wanous, 1992;Wanous and Reichers, 2000).
An additional difference is that orientation tends to occur when
a person joins an organization, whereas training could happen
at any stage during ones organizational tenure (Wanous and Re-
ichers, 2000). Job training for software professionals is a well-
researched area; previous studies have focused on different types
of roles in software such as developers (Johnson and Senges, 2010),
testers (Pham et al., 2015), volunteers in open source communities
(Panichella, 2015;Canfora et al., 2012;Sarma et al., 2016), and soft-
ware security (Papanikolaou et al., 2011). These studies highlight
the importance of training newcomers and suggest tools and best
practices (Panichella, 2015;Cherry et al., 2004;Sarma et al., 2016).
One of Van Maanen (1978)s six dimensions of organizational
socialization is the level of formality (formal vs. informal). A for-
mal socialization process is one in which newcomers are separated
from other employees as they are subjected to a program specifi-
cally tailored to them. An informal process does not differentiate
newcomers, and in such cases newcomers are expected to learn
on the job.
Many studies highlight a prevalent skill gap of newly hired
employees who have recently graduated (Byrne and Moore, 1997;
McGuire and Randall, 1998;Lethbridge, 1998;Lethbridge, 2000;
Tang et al., 2001;Brechner, 2003;Surakka, 2007;Tesch et al.,
2008;Lee and Fang, 2008;Simmons and Simmons, 2010;Rader-
macher and Walia, 2013;Pham et al., 2017). Technical support,
software installation, information management, and maintenance
of computer devices or components are the tasks performed by
most IT graduates (Legier et al., 2013). Three broad categories of
training can be identified: (1) practice-based learning (PBL) (John-
son and Senges, 2010), (2) class-based learning, and (3) mentoring
(Casado-Lumbreras et al., 2011), which are briefly summarized be-
low. Other training techniques include online training, tool based
training (Panichella, 2015;Cherry et al., 2004), and task curation
(Sarma et al., 2016). Organizations may follow one or a combina-
tion of these practices to train their newly recruited software pro-
fessionals.
PBL is a work-centered learning methodology (Johnson and
Senges, 2010), which is rooted in Lave and Wenger (1991)s general
theory of legitimate peripheral participation, which attempts to
create an environment conducive to growth and innovation within
the organization. Brown and Duguid (2000) described practice-
based learning as creating organizational conditions where new-
comers learn techniques of software development practices by
watching their fellow colleagues. It is carefully integrated as part
of the normal job of an employee and thus rendered invisible. In
terms of Van Maanen (1978)s dimensions of socialization, this cor-
responds to individual and informal organizational socialization.
Mentoring, involving connecting a newcomer (the mentee)
and a more experienced senior colleague (the mentor), is one of the
most common ways of transferringknowledge (Casado-Lumbreras
et al., 2011). The relationship between the mentor and the mentee
has been identified as one of the most important relationship in
a persons professional career. Apart from transferring skills and
knowledge, mentors also provide moral support to their protégés.
Mentors have a two-fold responsibility towards their mentees: ca-
reer development and psycho-social support. Career development
involves accustoming the newcomer towardst he necessary knowl-
edge and skills to succeed in the job. This is also referred to as
cross pollination (Bauer, 2010). This includes, for example, trans-
ferring knowledge about a programming language, framework, or
methodology. This can be done through one-on-one interactive
sessions, coaching, providing exposure, or giving challenging as-
signments. Psycho-social modeling is the informal aspect of men-
torship. It involves being a role model from whom the mentee can
seek personal guidance, acceptance, counseling, and friendship.
Fagerholm et al. (2014a) found that active mentoring of new
developers correlated with higher levels of activity, suggesting a
higher level of productivity. Based on an interview study with soft-
ware engineers, Enes (2005) found that this technique was more
successful than classroom-based teaching, with respondents indi-
cating that formal and organized teaching courses did not provide
adequate application-domain knowledge.
Whatever the training mechanisms, the main aim of training
is to ensure that a newcomer can perform the tasks of the job he or
she is recruited to do. We suggest that training is positively linked
to onboarding success, as defined above. Hence, we hypothesize
that:
Hypothesis 2 (H2). Training programs for newly recruited software
professionals are positively associated with onboarding success.
2.1.3. Support
The transition that new employees undergo to become a fully
functioning employee is not a set of discrete steps. Rather, it is a
continuous process that starts with orientation and training, and is
achieved through a continuous system of providing support and
feedback to the newcomer. Although university curricula teach
students the basic principles and concepts of software engineer-
ing, SE is a field where newcomers must continuously learn new
skills and technologies, such as new programming paradigms, lan-
guages, and techniques (Begel and Simon, 2008b). Other skills in-
clude the capability to create and debug specifications, document-
ing code, understanding and following a software development
process, managing projects, and working within a team. These
are skill gaps that are usually overlooked in university curriculum
(Byrne and Moore, 1997;Tang et al., 2001;Surakka, 2007) and even
a company-trained newcomer may have difficulty in understand-
ing these concepts. Therefore, a support framework should be pro-
vided in the workplace, so that newcomers can discuss their chal-
lenges and doubts with seniors or colleagues without feeling em-
barrassed or weak. For example, if a newcomer is having difficulty
in understanding a piece of code written by others who have left
the organization, he or she should not feel embarrassed to ask a
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 5
colleague or supervisor for help. We argue that the presence of
support mechanisms, such as the availability of help, appreciation
for a lack of newcomers knowledge, feedback, and the ability to
discuss personal issues that might affect performance, will posi-
tively link to newcomers onboarding experience—or, what we call
onboarding success. This in turn may lead to a better adjustment to
the new setting and an increase in self-efficacy. Hence, we propose
the following hypothesis:
Hypothesis 3 (H3). Offering support mechanisms to newly recruited
software professionals is positively associated with onboarding suc-
cess.
2.2. Onboarding success and organizational fit
Thus far, we have focused on activities that organizations can
organize and provide to newcomers in order to help improve those
newcomers onboarding experience. If that experience is positive,
newcomers will feel comfortable, accepted, and confident to do
their job. These activities—onboarding, training, support—are usu-
ally provided in the first stage of employment at an organization.
In the longer term, organizations will be interested in whether
employees will stay with the organization. Studies have demon-
strated that an individuals socialization trajectory and becoming
part of the core group in a team takes time and effort. Studies of
open source communities suggest that only a small part of the pe-
ripheral group of newcomers, who have their performance recog-
nized, are valued and eventually successful in their roles (Duche-
neaut, 2005;Fang and Neufeld, 2009). Qureshi and Fang (2011)
suggested that the lead time for different newcomers to attain
a core status within the work group may vary. This is why on-
boarding is also referred to as organizational socialization (Bauer,
2010). Following previous studies that have linked effective on-
boarding to job satisfaction (Klein et al., 2006;Cable et al., 2013;
Lavigna, 2009;Snell, 2006), we propose the following hypothesis:
Hypothesis 4 (H4). Software professionals degree of onboarding suc-
cess is positively associated with job satisfaction.
Further, new employees who are adequately socialized and
have effective relations with their peers will feel more adapted to
their new job demands, have an improvedlevel of self-efficacy, and
have a stronger attachment to the organization leading to greater
organizational commitment (Bauer et al., 2007). Fisher (1985)
found that sixty percent of employees consider strained relation-
ships with their peers as the reason for failed onboarding. We
therefore argue that, besides job satisfaction, which refers to con-
tentment with the position, successful onboarding experience is
also linked to good relationships in the workplace. Hence, we hy-
pothesize:
Hypothesis 5 (H5). Software professionals degree of onboarding suc-
cess is positively associated with the quality of their workplace rela-
tionships.
We refer to these two characteristics of (1) being content with
the job (job satisfaction) and (2) having good relationships within
the workplace as an employees organizational fit. Together,
these characteristics reflect a persons fit with the job and his
or her fit within the social environment of the workplace.
2.3. Organizational fit and turnover intention
Turnover intention is defined as a conscious and deliberate
willingness to leave an organization (Tett and Meyer, 1993). Ajzen
(1991)s Theory of Planned Behavior suggests that people act ac-
cording to their intentions and perceptions of control over their be-
havior (Lenberg et al., 2017). Despite the very strong relationship
between intended behavior and actual behavior, it is worth noting
that this is not a perfect relationship. Lee and Mitchell (1994) sug-
gest that employees maynot follow through with their intention to
leave until a precipitating shock event occurs, such as a reorgani-
zation or being assigned a new manager. It is also likely that acting
upon intentions may rely on the availability of other concrete and
more exciting opportunities.
As briefly pointed out, numerous studies have demonstrated
a positive link between effective onboarding and job satisfaction
and a negative link to turnover intention (Klein et al., 2006;Cable
et al., 2013;Lavigna, 2009;Snell, 2006). Employee turnover has
been the topic of extensive research—Hom et al. (2017) present a
concise overview of one hundred years of research.
Much of the research on employee turnover consists of
long-term studies; that is, these studies provide insights as to
the turnover decisions of long-term employees of organizations
(Holtom et al., 2008). Further, this relationship is debatable when
it comes to IT professionals. For example, Gupta et al. (2018)s
large-scale survey of newcomers in five industrial sectors found
that turnover intention was the highest in the IT sector. Gupta et al.
(2018) also found that newcomers with a high level of self-efficacy
(an indicator of successful onboarding (Bauer, 2010)) showed a
higher level of turnover intention. The study suggested that em-
ployees with higher self-efficacy are more confident in their ability
to switch over to a different job.
Given that the IT sector regularly expresses concerns about a
shortage of talent and the high cost of recruiting new staff (in-
curred partly due to the productivity delay mentioned in Sec-
tion 1),1it is worthwhile investigating this relationship for soft-
ware developers. Hence, we posit the following two hypotheses,
linking a newcomers organizational fit to a reduced intention to
leave his or her organization:
Hypothesis 6 (H6). Job satisfaction is negatively associated with
turnover intention.
Hypothesis 7 (H7). Workplace relationship quality is negatively as-
sociated with turnover intention.
Professionals who have recently started with an organization
are unlikely to have any intention to leave that organization within
a very short time. We argue that the perceived onboarding expe-
rience will not immediately correlate to turnover intention, but,
instead, that there are long-term mechanisms at work, specifically
job satisfaction and workplace relationship quality, two character-
istics of what we have labeled organizational fit. Job satisfaction
is not a state of being that appears immediately after joining an
organization—it is a sense of comfort and happiness that emerges
over time. Likewise, workplace relationships (i.e. relationships
with peers, managers, and friendships with colleagues) do not
form immediately, but rather developover time. Thus, job satisfac-
tion and workplace relationship quality are longer-term phenom-
ena, which we argue in this study, subsequently negatively corre-
late with an intention to leave the organization. In other words,
these two factors mediate the relationship between onboarding
success and turnover intention. Hence, we propose the following
hypothesis:
Hypothesis 8 (H8). Job satisfaction and workplace relationship
quality mediate the relationship between onboarding success and
turnover intention.
Figure 1 presents the full theoretical model (we note that there
is no standard notation for mediated relationships, hence the dot-
ted box and line for H8, which refers to the mediating role of job
1Reports appear regularly in the news, for example: https:
//www.forbes.com/sites/forbesbusinessdevelopmentcouncil/2018/06/29/
the-real- problem-with-tech-professionals-high- turnover/#4738d0aa4201
6G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
Training Onboarding
Success
Job
Satisfaction
Workplace
Relationship
Quality
Support
Orientation
Turnover
Intention
H1
H2
H3
H4 H6
H8
H5 H7
Organizational FitOnboarding activities
Fig. 1. Research Model
satisfaction and workplace relationship quality on turnover inten-
tion).
3. Research Design
In order to evaluate our theoretical model, we conducted a
cross-sectional survey, targeting software professionals with any
length of experience. We conducted a cross-sectional survey
rather than a survey within one specific organization, as this would
more likely provide the requisite variety in responses that is nec-
essary to evaluate a theoretical model such as ours. We selected
Partial-Least Squares Structural Equation Modeling (PLS-SEM) to
analyze the data. PLS-SEM is suitable to conduct exploratory
theory-development studies (Hair et al., 2011;Hair et al., 2016;
Russo and Stol, 2019). A well known alternative SEM approach is
covariance-based SEM (CB-SEM), which is more suitable for con-
firmatory research and tends to require larger sample sizes (Hair
et al., 2016). The remainder of this section proceeds as follows. Of
particular importance is that the constructs used in the hypotheses
are well defined and operationalized. Hence, we first discuss the
measurement model. We then discuss data collection and analysis
procedures.
3.1. Measurement Model
The theoretical model comprising the eight hypotheses are
based on a number of constructs, or so-called latent variables. A la-
tent variable cannot be directly measured or observed, but instead
is measured through a set of indicators or manifest variables. In our
model, all constructs are reflective (as opposed to formative).
Any change in a reflective construct is said to be reflected in its
indicators. That is, if the construct changes (which cannot be di-
rectly measured or observed), it will cause changes in its indica-
tors, which can be measured.
Defining the constructs of studies such as ours is particularly
important given their latent (unobservable) nature, and links di-
rectly to the issue of construct validity, which is concerned with
the question: does the researcher measure what she intends to
measure? A potential issue is that different studies may opera-
tionalize a certain construct differently by defining different indi-
cators. Further, particular care must be given to the issue of con-
struct clarity, so as to be able to clearly define and distinguish re-
lated, but different constructs.
We define the constructs of our model below, indicating what
we mean by each construct, and through which indicators we mea-
sured them. We adopted and tailored as needed existing measure-
ment instruments that have previously been used and validated.
Each construct had between two and six indicators, resulting in a
survey instrument of 30 questions. All indicators (questions) were
measured using a five-point Likert scale (ranging from 1=Strongly
Disagree to 5=Strongly Agree). The complete survey instrument is
available in Appendix A; below we summarize the origins of the
instruments to measure each construct.
•Orientation. Orientation comprises activities organized by an
organization that is specifically targeted at newcomers and usu-
ally within a relatively short time frame after newcomers en-
try to the organization. Orientation was measured using six
items adopted from Louis et al. (1983) and Wanous and Reich-
ers (2000). Items included attendance of orientation programs,
awareness of organizational rules and policies, and assignment
of a buddy or mentor, and items related to the organization or-
ganizing icebreaker events to facilitate meeting new colleagues,
as well as team activities.
•Training. Training is concerned with professionals task per-
formance, and thus this refers to specific activities to ensure
that newcomers can perform their tasks. We developed a new
instrument to measure software professionals training experi-
ence with four items based on prior literature. We captured for-
mal training for their job role (either a classroom based training
(Casado-Lumbreras et al., 2011) or one-on-one training from a
senior/mentor (Panichella, 2015)). Training on internal systems
and operating practices was adopted from Gupta et al. (2018)
without tailoring as it was directly applicable to software pro-
fessionals. The remaining two items were developed targeting
specifically software professionals, considering training in spe-
cific tools and methods, and the availability of a point of contact
or portal during training.
•Support. Support refers to the extent to which an organization
helps newcomers in the onboarding process. We derived an in-
strument with four items based on work by Bauer (2010) and
Gupta et al. (2018). The items measure the availability of a se-
nior or mentor to ask for help when the newcomer is stuck with
a given task, the extent to which newcomers feel weak or embar-
rassed to ask for help, the extent to which supervisors provide
constructive feedback, and the extent to which newcomers feel
that they can discuss personal issues when these affect their per-
formance.
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 7
•Onboarding Success. We define onboarding success as the ex-
tent to which a newcomer feels they are comfortable in their
new position. Bauer (2010) defines short-term outcomes that
reflect a successful onboarding experience: knowledge of orga-
nizational culture, role clarity, self-efficacy, and social integra-
tion; each of these is an item in our instrument. Further, Bauer
et al. (2007) suggest that role ambiguity (as opposed to role clar-
ity) is a source of dissatisfaction associated with stress, and in a
more general sense, joining a new workplace comes with a cer-
tain level of stress (Bourne, 1967;VanMaanen and Schein, 1979).
Hence, we captured the absence of stress as a fifth item to reflect
the construct onboarding success as defined above.
•Job Satisfaction. Job satisfaction has been studied extensively
in the management literature. Spector (1985) developed an in-
strument of 36 questions to measure job satisfaction. We found
that some of the questions were overlapping, so we selected six
indicators that are appropriate indicators of the construct job
satisfaction.
•Workplace Relationship Quality. We define workplace rela-
tionship quality as a persons perceived quality of his or her re-
lationships within the workplace. We developed an instrument
with three items to measure this construct. Sias (2005) studied
two types of workplace relationships: supervisor-subordinate
and peer co-worker relationships. We defined one indicator
for each of these types of relationships for the workplace rela-
tionship quality construct. Korte and Lin (2013) suggested that
newcomers not only seek acceptance into the group, but also
friendship (affect). Several others have studied friendships in
the workplace (Sias and Cahill, 2013;Rawlins, 1992). Hence, we
defined a third item to capture workplace friendships.
•Turnover Intention. Turnover intention is defined as a respon-
dents inclination to leave their organization. We adopted a two-
item instrument from Metcalf et al. (2015), which queries (1)
whether respondents think of quitting their job frequently, and
(2) whether respondents plan to search for a new job within the
next 12 months.
3.2. Data collection and analysis
We implemented our survey instrument with SurveyMonkey,
and advertised it through a number of channels. We distributed
the survey through our professional network; contacts were also
invited to share the link with their colleagues. We also shared the
link on Twitter; this microblogging platform provides analytical
data about the tweet, indicating the tweet was retweeted 20 times,
and the link was clicked 30 times. The total number of engage-
ments (a metric defined by Twitter) was 72—this is the number
of Twitter users that the tweet was presented to. While the survey
was in principle anonymous, we did offer respondents an option
to enter their email address voluntarily if they were interested in
receiving the results of the study; a number of respondents were
subsequently sent a preliminary report with some of the findings.
We did not capture any geographical data or information about re-
spondents employers. We received 102 complete responses that
could be used for analysis. While sample size is commonly an is-
sue for covariance-based SEM (CB-SEM) due to the requirements
for its computation, PLS can also be used with more modest sample
sizes. A common minimum threshold for a sample is the so-called
10 x rule, which states that the sample should be a minimum of
ten times the largest number of structural paths directed at a latent
construct (Hair et al., 2011)—in our study, the maximum number
of structural paths to a latent construct is three, suggesting a sam-
ple size of only 30. This rule of thumb has been criticized, in par-
ticular when such calculations leads to small sample size recom-
mendations; our sample size is more than three times this number.
Table 2
Respondents roles
Role Count
Software developer/engineer 41
Analyst 16
Software tester 16
Technical support 6
Technical consultant 6
Researcher 6
Project manager 4
Web designer 4
Other 19
Table 3
Respondents total work experience, and experience in current
role
0-3 yr. 3-7 yr. 7+ yr.
Total experience 59 20 23
Experience in current
role
78 19 5
Within the software engineering domain, studies using PLS-SEM
commonly use samples between 50 and 100 (Parolia et al., 2013;
Parolia et al., 2015;Vijayasarathy and Turk, 2012).
We also conducted a power analysis, using the freely available
tool G*Power (Faul et al., 2009).2Given the exploratory nature of
this study, we used the threshold value for a medium effect size
(0.15 (Cohen, 1988)), a significance level of 0.05, and a default value
for the power (1-β) of 0.8 (Marcoulides and Saunders, 2006). The
maximum number of predictors is 3 in our model. This calculation
indicated a minimum sample size of 77, well below our sample of
102. Using a higher value for power of 0.9 yielded a minimum sam-
ple of 99.
Table 2 presents aggregate information about the respondents
roles, and Table 3 presents the total number of years of work ex-
perience, and the number of years in their current role. Most re-
spondents identified as a software developer/engineer, analyst, or
software tester. A small number identified as a project manager or
director. Not all respondents provided this information. In terms
of experience, 59 respondents (58%) indicated they had 0-3 years
total work experience, but 78 reported to be in their current role
only for 0-3 years. Almost all respondents were in their current
role for up to 7 years; only 5 reported to be in the current role for
7+ years.
A variety of PLS-SEM software packages is available; we used
the software package SmartPLS version 3.2.8 for the analyses, the
results of which are presented in Section 4. The analysis procedure
for PLS-SEM consists of two main steps, with several tests and pro-
cedures in each step. The first step is to evaluate the measurement
model, which empirically assess the relationships between the in-
dicators and the constructs. The results of this step are presented
in Section 4.1. The second step is to evaluate the theoretical or struc-
tural model which represents the set of hypotheses—thus, at is in
this step that the hypotheses are evaluated. We present the results
of the second step in Section 4.2.
4. Results
4.1. Evaluation of the measurement model
Before the structural model can be evaluated, we evaluate the
measurement model. We discuss the internal consistency reliabil-
ity, convergent validity, and discriminant validity.
2G*Power has a wide range of tests; we used the following settings: Test
family: F-tests; statistical test: linear multiple regression with fixed model, R2
deviation from zero; type of power analysis: a priori analysis.
8G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
Table 4
Internal consistency reliability
Construct Cronbach αComposite Reliability Average Variance Extracted (AVE)
Orientation 0.709 0.882 0.602
Training 0.761 0.847 0.583
Support 0.716 0.823 0.539
Onboarding success 0.774 0.846 0.525
Workplace relationship quality 0.738 0.848 0.651
Job Satisfaction 0.834 0.882 0.602
Turnover intention 0.714 0.871 0.772
Table 5
Cross loadings of the retained indicators on the constructs (A complete list of the items is available in Appendix A)
Item Item Description Orientation Training Support Onboarding
Success
Job
Satisfaction
Social
Acceptance
Turnover
Intention
or_3 Buddy/mentor assigned to help 0.730 0.345 0.398 0.334 0.462 0.157 –0.205
or_4 Made aware of challenges 0.660 0.492 0.205 0.264 0.302 0.203 –0.136
or_5 Icebreakers to meet colleagues 0.724 0.494 0.144 0.315 0.272 0.219 –0.150
or_6 Org. has team days / activities 0.784 0.282 0.307 0.470 0.484 0.505 –0.258
tr_1 Formal training program 0.269 0.764 0.112 0.211 0.194 0.114 –0.094
tr_2 Internal system training 0.423 0.808 0.200 0.276 0.370 0.096 –0.188
tr_3 Training in technology, methods 0.434 0.809 0.139 0.280 0.307 �0.008 –0.195
tr_4 Point of contact / online portal 0.431 0.664 0.144 0.313 0.294 0.104 0.124
su_1 Can seek help if stuck 0.264 0.241 0.662 0.444 0.383 0.270 –0.124
su_2 Not embarrassed asking for help 0.217 0.089 0.741 0.441 0.356 0.292 –0.270
su_3 Supervisor ongoing feedback 0.185 0.155 0.742 0.482 0.379 0.256 –0.133
su_4 Speak to supervisor if personal issues affect perfor-
mance
0.392 0.113 0.787 0.609 0.654 0.549 –0.302
os_1 Joining new workplace less stressful 0.346 0.464 0.430 0.662 0.455 0.246 –0.032
os_2 Familiarity with organizations culture 0.388 0.305 0.336 0.659 0.526 0.176 –0.137
os_3 Understand roles responsibilities 0.336 0.332 0.534 0.730 0.564 0.447 –0.270
os_4 Confidence in capability to do job 0.332 0.110 0.497 0.749 0.463 0.380 –0.255
os_5 Socially integrated in workplace 0.397 0.165 0.622 0.811 0.546 0.587 –0.315
js_1 Fair chance on promotion 0.487 0.320 0.498 0.626 0.845 0.358 –0.395
js_2 Org. offers growth opportunities 0.398 0.380 0.490 0.626 0.845 0.358 –0.395
js_3 Fair compensation 0.376 0.348 0.318 0.441 0.692 0.278 –0.102
js_4 Achievements recognized 0.337 0.203 0.461 0.478 0.763 0.315 –0.328
js_6 I am satisfied with my job 0.474 0.297 0.603 0.603 0.858 0.409 –0.438
rq_1 Professional relations with peers 0.348 0.188 0.416 0.532 0.435 0.871 –0.191
rq_2 Friends in the workplace 0.245 0.029 0.199 0.305 0.209 0.759 –0.081
rq_3 Professional relations with seniors 0.368 –0.014 0.498 0.408 0.411 0.785 –0.254
ti_1 I frequently think of quitting –0.300 –0.106 –0.323 –0.345 –0.407 –0.193 0.923
ti_2 Will look for jobs within next year –0.148 –0.078 –0.160 –0.144 –0.271 –0.219 0.832
4.1.1. Internal consistency reliability
Internal consistency reliability assesses how well the different
indicators are able to measure the constructs reliably and consis-
tently. There are several tests to measure this. We performed the
Cronbachs alpha and Composite Reliability tests. Cronbachs al-
pha tests generally show lower values of reliability and are more
conservative compared to composite reliability, which sometimes
overestimates the values (Hair et al., 2016). The true measure of
internal consistency reliability lies between the lower bound of
Cronbachs alpha and upper bound of composite reliability. For
exploratory research, values of 0.6-0.7 are acceptable, while for re-
search in a more advanced stage values between 0.7 and 0.9 are
recommended (Hair et al., 2016). Values below 0.6 suggest a lack of
internal consistency reliability, whereas values over 0.95 suggest
that indicators are too similar and therefore not desirable. Table 4
shows that the Cronbach alpha and CR for all fall within the range
0.7-0.9.
4.1.2. Convergent validity
Convergent validity measures how well different indicators of
a construct correlate positively with one another. All constructs
in our model are reflective (not formative), which means that in-
dicators are considered to be different ways to measure the same
construct—they should share a considerable proportion of vari-
ance, which means that they converge. To assess convergent va-
lidity, two metrics are important: the Average Variance Extracted
(AVE), and the outer loadings of a constructs indicators. The AVE
values should be at least 0.5, and Table 4 shows that all AVE values
are all above that threshold.
Table 5 reports the outer loadings of all items. A standard rule
of thumb is that outer loadings should be higher than 0.70 (Hair
et al., 2016), but this does not mean that values below this thresh-
old should always be removed. If outer loadings are between
0.40 and 0.70, the effect of removing them on the AVE should be
considered—if the AVE improves significantly, the items should be
removed, but otherwise they can be retained. In this case, we
found that AVE values of all the constructs, except orientation,
were above the desired threshold of 0.50. Hence, two indicators
of the orientation construct were removed, leading to an improve-
ment of the AVE for orientation from 0.416 to 0.527. After their re-
moval, all outer loadings were above 0.65 except one (js_5, impor-
tance of work), which had a negativeef fecton discriminant validity
(discussed below); hence, we also removed this indicator, leaving
all outer loadings with values higher than 0.659.
4.1.3. Discriminant validity
Discriminant validity refers to the extent to which the differ-
ent constructs capture different phenomena or concepts. In other
words, it is a measure of how distinct each construct is in relation to
other constructs. It implies that each construct is unique and rep-
resents characteristics that are not measured by other constructs.
There are three common ways to assess discriminant validity. First,
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 9
Table 6
Fornell-Larcker criterion: correlations among the constructs and square roots of the AVE values (on diagonal)
Construct 1 2 3 4 5 6 7
1. Orientation 0.726
2. Training 0.526 0.763
3. Support 0.371 0.199 0.734
4. Onboarding success 0.494 0.363 0.683 0.724
5. Workplace relationship quality 0.406 0.099 0.485 0.535 0.807
6. Job Satisfaction 0.539 0.392 0.623 0.705 0.460 0.776
7. Turnover intention –0.269 –0.107 –0.289 –0.297 –0.230 –0.396 0.879
Table 7
Heterotrait Monotrait (HTMT) ratios
Construct 1 2 3 4 5 6 7
1. Orientation
2. Training 0.731
3. Support 0.503 0.271
4. Onboarding success 0.644 0.483 0.884
5. Workplace relationship quality 0.500 0.199 0.603 0.641
6. Job Satisfaction 0.671 0.487 0.765 0.869 0.552
7. Turnover intention 0.342 0.273 0.372 0.366 0.304 0.466
we investigated the cross-loadings: the outer loadings of a con-
structs indicators should be higher for that construct than on any
of the other constructs. If an indicator would load higher onto a dif-
ferent construct than the one that it purportedly measures, then
that suggests the indicator is a better measure for that other con-
struct. Table 5 shows that (inspecting row by row) all constructs
indicators load highest onto their respective constructs.
The second approach to assess discriminant validity is evalu-
ating the Fornell-Larcker criterion. This criterion states that the
square root of a constructs AVE should be higher than that con-
structs correlation with other constructs. In plain terms, a con-
struct should share more variance with its own indicators than
with other constructs. We observed that the AVE value for on-
boarding success was slightly lower (approx. 0.01) than the cor-
relation with job satisfaction; we resolved this by removing item
js_5 as mentioned above. Table 6 lists the construct correlations,
with the square roots of the AVE values on the diagonal. All square
roots of the AVE values comply with the Fornell-Larcker criterion.
Third, we also considered Henselers Heterotrait-Monotrait
(HTMT) ratio (Henseler et al., 2015) (see Table 7). The cut-off
value is 0.9 beyond which discriminant validity could be consid-
ered problematic (Henseler et al., 2015), though some researchers
recommend a more conservative cut-off of 0.85 (Hair et al., 2016).
Table 7 shows that most HTMT ratios are below 0.85, with only two
ratios between 0.85 and 0.9 (onboarding success/support, and job
satisfaction/support). Besides these cut-off values, HTMT ratios
should also be significantly different from 1.0; this can be calcu-
lated using a bootstrap procedure, which calculates bias-corrected
confidence intervals for all ratios. (We discuss the bootstrap proce-
dure in more detail below.) None of these included the value 1.0,
indicating that all HTMT ratios were acceptable. Based on these
three tests to assess discriminant validity, we conclude that the
discriminant validity of our study is satisfactory.
4.2. Evaluation of the structural model
We now turn our attention to the evaluation of the structural
model, which includes the evaluation of the hypotheses.
4.2.1. Assessing collinearity
Our theoretical model consists of seven constructs, of which
three are exogenous (orientation, training, and support). To en-
sure that the exogenous constructs are independent, we evaluate
their collinearity by means of the Variance Inflation Factor (VIF).
A widely accepted cut-off value for the VIF is 5 (Hair et al., 2011),
though collinearity issues may also occur with VIF values between
3 and 5, which is why Hair et al. (2019) recommend a cut-off value
of 3. In our model, all VIF values are below this more conserva-
tive cut-off value between 1.1 and 2.3, indicating that there are no
collinearity issues in our model.
4.2.2. Path coefficients and significance
PLS does not make any assumptions about the distribution
(such as a Normal distribution) of the data; therefore, it cannot use
any parametric testsof significance. In order to determine whether
path coefficients are statistically significant, PLS packages depend
on a bootstrapping procedure. This involves drawing a large num-
ber (typically five thousand) of random subsamples with replace-
ment; replacement is needed to ensure that all subsamples contain
the same number of observations as the original data set. For each
subsample, the PLS path model is estimated; these sets of coeffi-
cients form a bootstrap distribution, which can be considered as an
approximation of the sampling distribution (following the Central
Limit Theorem). From this, a standard error and standard devia-
tion are determined (Hair et al., 2016), which can subsequently be
used to make statistical inferences. Table 10 shows the results for
our eight hypotheses, including the mean of the bootstrap distribu-
tion, the standard deviation, the 95% confidence interval, and the p
values. Based on these results, we found support for Hypotheses 1,
and 3-6. We only found weak support for Hypothesis 2 (p=0.064).
Hypothesis 7 was not supported (p=0.663).
Hypothesis 8 proposed that job satisfaction and workplace re-
lationship quality mediate the link between onboarding success
and turnover intention. To evaluate mediating relationships, we
must compare the indirect paths suggested by the mediators, to
the direct paths (Zhao et al., 2010;Nitzl et al., 2016). Variables may
have no mediating effect (indirect effect is insignificant), a partial
mediating effect (if the direct effect is significant), or a full mediat-
ing effect (if the direct effect is insignificant).
Table 2 shows that the direct association between onboard-
ing success and turnover intention is not significant (p=0.954).
On the other hand, the indirect association between onboarding
success and turnover intention with job satisfaction as a medi-
ator was statistically significant (p<0.02). The association be-
tween onboarding success and turnover intention mediated by
workplace relationship quality was not statistically significant
(p=0.673). These findings lend support to our hypothesis that job
10 G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
Training Onboarding
Success
Job
Satisfaction
Workplace
Relationship
Quality
Support
Orientation
0.140*
0.579**
0.705**
0.535**
Turnover
Intention
or_3
or_4
or_5
tr_1
tr_2
tr_3
tr_4
su_1
su_2
su_3
su_4
ti_1
ti_2
os_1 os_2 os_3 os_4
rq_1 rq_2 rq_3
js_1 js_2 js_3 js_4 js_6
0.730
0.660
0.206** -0.363**
0.809
0.764
0.808
0.664
0.662
0.741
0.742
0.787
0.845
or_6
0.724
os_5
0.706 0.692 0. 763 0.858
0.6590.661 0.8110.749
0.730
0.784
0.871 0.759 0.785
0.923
0.832
-0.010 (n.s.)
-0.058 (n.s.)
R2=0.547
R2=0.497
R2=0.286 R2=0.160
Fig. 2. Outer loadings (measurement model) and path coefficients (structural model) (* p<0.10 indicated by a dashed line, ** p<0.05). Non-significant
links are indicated with a dotted line.
Table 8
Coefficients of determination
Construct R2Q2
Onboarding success 0.547 0.244
Job satisfaction 0.497 0.266
Workplace relationship quality 0.286 0.165
Turnover intention 0.160 0.089
Table 9
Effect sizes for all constructs (with specified endogenous
variable to which the effect size applies)
Construct f2
Orientation 0.061
Training 0.031
Support 0.638
Onboarding success
Job satisfaction 0.989
Workplace relationship quality 0.401
Turnover intention 0.000
Workplace relationship quality 0.003
Job satisfaction 0.077
satisfaction fully mediates the relationship between onboarding
success and turnover intention, but workplace relationship qual-
ity does not. Section 5 presents a discussion of these results.
4.2.3. Coefficient of determination and effect sizes
This stage of analysis helps assess the relationship between
constructs and the predictive capabilities of the model. The R2val-
ues of the four endogenous variables in our model are listed in Ta-
ble 8.R2values range between 0 to 1, with values of 0.75, 0.50, or
0.25 considered substantial, moderate, or weak respectively (Hair
et al., 2016;Henseler et al., 2009). We found moderate values for
onboarding success and job satisfaction, but only a weak result for
workplace relationship quality and turnover intention.
The value of R2will change when an exogenous construct is
removed from the model—a measure for the extent to which the
exogenous construct contributes to an endogenous constructs R2
value is the f2effect size. Table 9 lists the f2values for each of
the constructs. An effect size below 0.02 means that there is no ef-
fect of the exogenous construct on the endogenous construct. The
threshold values 0.02, 0.15, and 0.35 refer to, respectively, small,
medium, and large effects (Cohen, 1988). The table shows that of
the three constructs associated with onboarding success, support
makes the largest contribution to the R2with an f2effect size of
approximately 0.64. Orientation and training have rather small ef-
fects with effect sizes of 0.03 and 0.06, respectively.
Finally, we also inspected Stone-Geissers Q2value which is
an indicator of the models predictive relevance (Hair et al., 2016)
(these can be obtained through a blindfolding procedure; Hair et al.
discuss this in detail). Q2values are calculated only for reflective
endogenous constructs—a value larger than 0 indicates the con-
struct has predictive relevance. The same thresholds as for the R2
apply to Q2(0.02. 0.15, 0.35). Table 8 shows that all Q2values are
greater than zero, suggesting our model has predictive relevance
as well.
5. Discussion and conclusion
5.1. Discussion of results
In this study, we have drawn on the wider literature beyond
software engineering, including management and psychology lit-
erature, to derive a theoretical model for studying onboarding suc-
cess of software professionals, and its relation to their inclination
to leave their job. Evaluation of the theoretical model helps us un-
derstand the role of a number of activities that have been shown to
be important in onboarding in general but not yet within the soft-
ware development domain. Our analysis highlights a number of
key findings and implications—Table 11 provides a summary.
We found support for H1 which proposed a positive associa-
tion between orientation and onboarding success. Orientation typ-
ically happens during the first days or weeks of an employee join-
ing an organization, and involves giving out essential information
about the rules and policies of the company and helping newcomer
interaction. Of the six indicators, we retained four after inspecting
the outer loadings and the AVE. The items that were removed re-
flected the extent to which respondents attended an orientation
program, and whether they were made aware of organizational
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 11
Table 10
Path coefficients, bootstrap estimates, standard deviations, confidence intervals, and p values
Hypothesis Coefficient Bootstrap mean Std.dev. 95% CI p
H1: Orientation →Onboarding success 0.205 0.211 0.080 (0.024, 0.347) 0.010
H2: Training →Onboarding success 0.140 0.156 0.076 (–0.029, 0.265) 0.064
H3: Support →Onboarding success 0.579 0.576 0.068 (0.439, 0.702) 0.000
H4: Onboarding success →Job satisfaction 0.705 0.706 0.062 (0.548, 0.801) 0.000
H5: Onboarding success →Workplace relationship quality 0.535 0.546 0.079 (0.354, 0.671) 0.000
H6: Job satisfaction →Turnover intention –0.363 –0.366 0.140 (–0.591, –0.020) 0.010
H7: Workplace relationship quality →Turnover intention –0.058 –0.055 0.133 (–0.306, 0.223) 0.663
Mediators
H8:
Direct effect:
Onboarding success →Turnover intention –0.010 –0.014 0.170 (–0.325, 0.344) 0.954
Indirect effects:
Onboarding success →Job satisfaction →Turnover Intention –0.256 –0.257 0.062 (–0.437, –0.026) 0.017
Onboarding success →Workplace relationship quality →Turnover Intention –0.031 –0.028 0.073 (–0.170, 0.125) 0.673
Table 11
Summary of results and implications
Hypothesis Findings Implications
H1: Orientation →
onboarding success
Supported. While the standardized path
coefficient was moderate (0.2), the orientation
program does contribute to onboarding
success, though the effect size is low (0.06).
Orientation activities have a moderate correlation with onboarding success.
Organizations should leverage the opportunity to give newcomers a good
introduction to help people settle in.
H2: Training →
onboarding success
Weak support. Training has a low standardized
path coefficient (0.14) with a pvalue of 0.06,
meaning it is not statistically significant
following the standard alpha level of 0.05.
Small effect size (0.03).
Training does not seem to be helpful towards successfully onboarding new
recruits. It is likely that the knowledge required for a given job is highly
specific, and too much to cover during a short-term formalized training
program. Organizations might do better through catering for on-the-job
training of new recruits.
H3: Support →
onboarding success
Supported. Support was found to be the largest
and most significant factor associated with
onboarding success, with a standardized path
coefficient of 0.58, and an effect size of 0.64.
Providing ongoing support to new recruits so that they feel supported in their
job is likely to be most important in getting new staff to settle in.
Organizations should create an environment where people feel supported
and safe to ask for help.
H4: Onboarding Success
→Job Satisfaction
Supported. Onboarding success has a
considerable (standardized path coefficident of
0.7) and statistically significant positive
association with job satisfaction.
Ensuring the onboarding process is successful may be key to achieving a high
level of job satisfaction. Easing newcomers into the new job so as to
integrate them is key when designing onboarding programs.
H5: Onboarding Success
→Workplace
Relationship Quality
Supported. Onboarding success has a
considerable (standardized path coefficient of
over 0.5) and statistically significant positive
association with workplace relationship
quality.
Professionals who perceive their onboarding experience to be successful also
have good relationships within the workplace. They fit in socially, which
may reduce conflicts in the workplace, which should be of interest to
organizations; however, studies are needed to explore this further.
H6: Job Satisfaction →
Turnover intention
Supported. Job satisfaction has a statistically
significant and considerable (standardized
path coefficient of −0.36) inverse association
with turnover intention. Hence, software
professionals who are content in their job are
less likely to leave the organization.
This study finds support for this hypothesis in the software domain. Given the
high cost of turnover in this domain, it may be of interest to conduct studies
that explore what dissatisfies professionals so that organizations can take
countermeasures in order to ensure that job satisfaction levels remain high.
H7: Workplace
relationship quality →
Turnover intention
Not supported. Having good relationships with
peers and managers does not associate with a
lower intention to leave the organization
(standardized path coefficient <0.06, p>0.6).
While having good relationships within the workplace, perhaps even having
friends, may be good for productivity but it bears no effect on a
professional’s intention to stay with or leave the organization. Organizations
may still want to take measures to improve workplace relationships for other
reasons (e.g. productivity), but it does not help to retain staff.
H8: Job satisfaction and
workplace
relationship quality
mediate onboarding
success →turnover
intention
Partially supported. Job satisfaction fully
mediates the relationship between onboarding
success and turnover intention (standardized
path coefficient −0.25, p= 0.017), but
workplace relationship quality does not
(−0.031, p= 0.67). No direct effect from
onboarding success to turnover intention
(−0.01, p= 0.95).
Job satisfaction plays a key role in achieving a short-term organizational
socialization (onboarding) and longer-term retaining of staff (i.e. low
turnover). Having good workplace relationships does not. Organizations that
wish to focus on retaining staff should take measures to naturalize new
recruits as well as take measures that staff are content with their job in the
longer run. Social activities may help reduce friction in day-to-day conflicts,
this will not help in retaining staff.
rules and policies. The retained indicators all refer to social as-
pects, including a buddy to help people settle in, dealing with chal-
lenges, and icebreaker activities as well as social gatherings; the
removed indicators both referred to more objective (rather than
opinionated) aspects, such as awareness of rules and regulations.
We found only weak support for H2 (p<0.10), suggesting that
training of newcomers only has a marginal link to onboarding suc-
cess. Task-oriented training can have a direct effect on role clarity
and self-efficacy, which are indicators of onboarding success—this
in turn may help to address the skills gap mentioned in Sections 1
and 2. This might also explain training in internal systems and op-
erating practices (environment), and tools/technology used for the
job have the highest loadings on the training construct (both over
0.8). Our study has focused on a rather narrow meaning of train-
ing to facilitate a clear and precise definition and measurement.
Recently, Baltes and Diehl (2018) have presented a more holistic
treatment of the notion of developer expertise.
12 G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
H3 proposed a positive association between support and on-
boarding success—this hypothesis is also strongly supported by
our study. Support can be considered a more continuous process,
and ongoing support will make newcomers at ease when seeking
help from seniors and peers regarding professional and personal
issues, without invoking feelings of being judged or embarrassed.
We found that support provided to newcomers is the largest con-
tributor to onboarding success (standardized path coefficient of
ca. 0.58). The effect size (f2=0.638) analysis also shows that the
support construct makes a large contribution to the coefficient of
determination (R2) of onboarding success in our model. Over 80
percent of respondents felt they could ask for help from a senior
in matters related to the job tasks, and over 70 percent indicated
that they would not feel weak or embarrassed in doing so. This sug-
gests that an organizational culture that encourages transparency
and helping others can help newcomers to onboard successfully.
Newcomers initially mayhave many questions about their new job,
and having a personal point of contact to address these questions
will help to ease into the new work environment.
We found support for H4 and H5, which suggested that on-
boarding success (which tends to be a short-term outcome, as it
takes place within the first few months) is positively associated
with job satisfaction and workplace relationship quality. Both job
satisfaction and relationship quality are longer-term, effects, in
that both are emergent perceptions; people need time before they
can reflect on their relationship with their job (job satisfaction) and
relationships with colleagues (with peers, managers, and creating
friendships) also take time to shape. Although most respondents
showed high levels of job satisfaction, the level of satisfaction
with remuneration was relatively low compared to other indica-
tors. Eighty-six percent and 89 percent of respondents said they
had good personal and professional relationships with their col-
leagues, respectively. This might be a result of either the efforts of
the organization by organizing social events or through their uni-
versity social network—the latter is a possibility because almost
58 percent of the respondents are recent graduates with less than
three years of professional work experience (see Table 2).
Job satisfaction was found to have a negative link to turnover
intention (H6), implying that people who are happy with their job
are less likely to leave their employer. The quality of their relation-
ships within the workplace (H7), on the other hand, did not seem
to have a significant link to turnover intention: people may have
very good relationships with their peers, managers, and even have
friends within the workplace, this does not seem to stop people
from considering to leave their position.
Finally, we argued that job satisfaction and workplace relation-
ship quality mediate the relationship between onboarding success
and turnover intention (H8). We only found partial support for this.
We found support for the mediating role of job satisfaction, sug-
gesting that a successful onboarding experience (a short-term
experience, soon after recruitment) may help to achieve job satis-
faction (a longer-term state of contentment), which in turn may
help to retain staff. Workplace relationship quality, on the other
hand, is not a mediator; while we found that a successful onboard-
ing experience was positively associated with the quality of work-
place relationships (H5), this did not translate to a lower turnover
intention—having already established that no support was found
for H7, this is not surprising.
5.2. Limitations of this study
5.2.1. Construct validity
We adopted and tailored existing measurement instruments,
and developed derived measurement instruments for some con-
structs based on prior literature. Our analysis of the measurement
model confirmed that our constructs were internally consistent,
and scored satisfactory on convergent and discriminant validity
tests. We defined a new construct called workplace relationship
quality; we did not identify a construct in prior literature that rep-
resents the quality of relationships within a workplace. Though
newly defined, it scored well on the tests mentioned above.
5.2.2. Internal validity
This study is a sample study rather than an experimental study
(i.e., we made no interventions), and drawing causal relationships
is typically not possible (Stol and Fitzgerald, 2018). Our hypothe-
ses propose associations between different constructs rather than
causal relationships. While it is clear that activities such as orien-
tation and training tend to occur at an early stage of an employees
tenure and that job satisfaction tends to emerge over time (i.e.,
after orientation and training), we cannot exclude the possibility
that other factors are at play. Two of four coefficients of determi-
nation (R2) can be considered moderate at values over 0.5; thus,
while other factors are likely to play a role, these results represent
a useful starting point for future studies.
5.2.3. External validity
This survey was conducted online and anonymously, and thus
we cannot report any details on the extent to which our sample
was representative. The nature of our sample is a convenience
sample, which we contacted through our professional networks
and through social media. However, in our study we sought to get
responses from software professionals in general—we deem it un-
likely that people not active in the software industry would have
completed the survey. Table 2 shows that our population sam-
ple represents a variety of software professionals; hence we argue
that the sample served our study goal, namely to seek evidence for
our theoretical model focused on software professionals. The re-
sponses were sufficiently consistent to find full empirical support
(p<0.05) for five of eight hypotheses, weak support for one (p<
0.10), and partial support for another (H8). However, we suggest
further studies to replicate our findings.
5.3. Conclusion and future work
Designing a successful onboarding program is a key part of any
organizations talent management and retention strategy. As pre-
viously discussed, IT professionals show one of the highest levels
of turnover intention compared to employees in other industries.
While several field studies exist, the software engineering litera-
ture lacks theoretical models that help organizations understand
which factors play a role in achieving success in the onboarding
process, and how this might translate to longer-term organiza-
tional fit as manifested by job satisfaction and workplace relation-
ships, and ultimately, a reduced level of turnover.
This study sought to develop such a theoretical model specif-
ically targeting the software engineering domain, which may in-
form the development of onboarding processes for software pro-
fessionals. Drawing from the literature on onboarding from other
disciplines as well as studies of onboarding in the software en-
gineering literature, we derived a theoretical model comprising
eight hypotheses. We evaluated these hypotheses through a sam-
ple study, using an online survey instrument. Based on 102 re-
sponses, we found (partial) support for six hypotheses.
The strongest statistical significance was found forthe associa-
tion between support and onboarding success. Support is a contin-
uous rather than a discrete time-bound activity, and can be easily
overlooked; rather than pushing information and training onto
newcomers, new recruits pull support from senior colleagues
and are offered a safe environment to ask questions without evok-
ing a sense of embarrassment. Our study shows that support is
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 13
key to a newcomers organizational socialization (Bauer, 2010)
or naturalization (Sim and Holt, 1998). Software organizations
are increasingly growing and operating in ever more complex and
diverse settings. Technologies are evolving constantly, and new
methods and practices emerge continuously—with methods in use
becoming irrelevant. Few other industries see such continuous
change and evolution as the IT industry. This could be why an en-
vironment that provides constant support and feedback to its em-
ployees is indispensable to assist its employees in keeping up-to-
date. Orientation and training, on the other hand, had only modest
associations with onboarding success. Hence, one of the two most
important recommendations from this study is not only emphasiz-
ing the importance of a supporting environment, but also the rela-
tively insubstantial contribution of potentially expensive orienta-
tion and training programs to the success of onboarding programs.
Besides onboarding activities, our study also considers what
we have termed organizational fit of a new recruit, which refers
to how content that person is with the job, and how well the per-
son fits in socially. We found that successful onboarding experi-
ences correlate positively with both these aspects, and so a good
onboarding experience helps to establish a persons organizational
fit. Of the two aspects, only job satisfaction had a negativeassocia-
tion with turnover intention; the other aspect, a persons relation-
ships within the workplace, did not. Further, we found that job
satisfaction mediates the relationship between onboarding suc-
cess and turnover intention; whereas a successful onboarding ex-
perience does not associate with a lower turnover intention, an in-
direct association does exist through the job satisfaction construct.
Based on our findings, we suggest a number of avenues for fu-
ture research.
While job satisfaction seems to play a major role in a software
professionals decision to stay or leave an organization, the reality
in the IT sector faces significant challenges in staff retention. Fu-
ture work might explore which other factors play a role in a deci-
sion to leave an organization.
The results of our study emphasize the importance of support—
future work might explore effective ways that organizations can
offer this support to software professionals, which likely must be
tailored to different roles within the software industry.
The results of this study suggest a positive link between on-
boarding success and job satisfaction, but it is highly likely that this
relationship may change over time; that is, the sense of onboard-
ing success might fade, and other factors might start to impact
software professionals job satisfaction in more significant ways.
Whereas our study did not find any benefit of having good
workplace relationships (in terms of turnover intention), it is likely
that staff who have good relationships within the workplace are
happier, which some studies suggest will benefit productivity
(Graziotin et al., 2018). It is worth exploring how good workplace
relationships can be established (aside from having a successful
onboarding experience), and which effects this might have on soft-
ware professionals. In this context, it is also worth noting that our
study did not distinguish between internal and external turnover
intention; that is, poor workplace relationships may not encour-
age people to leave the organization (i.e., external turnover), but it
might encourage them to join a different team or unit within the
same organization (i.e., internal turnover).
Furthermore, future research could also investigate software
professionals character attributes, personality, attitudes, and be-
liefs in relation to the relationships they forge within the work-
place. It is likely that these factors affect how software profession-
als perceive the value of workplace relationships, and therefore it
is of interest to evaluate whether or not these different attitudes
and personal values relate to an intention toleave the organization.
The software industrys landscape is fast-moving, with many
start-ups, acquisitions, and mergers. When a company is acquired
by larger ones, its staff are effectively hired wholesale. Our model
has not considered this option.
To conclude, this study contributes to the relatively limited lit-
erature within the software engineering domain on onboarding
of software professionals. While considerable attention has been
paid to onboarding in open source communities in recent years,
most software that is developed remains to be closed source. It
is our hope that this study offers a good starting point for future
work.
Acknowledgment
We are grateful to the anonymous reviewers who offered gen-
erous and constructive feedback which has led to dramatic im-
provements of this manuscript. This work was supported, in part,
by Science Foundation Ireland grant 15/SIRG/3293 and 13/RC/2094
and co-funded under the European Regional Development Fund
through the Southern & Eastern Regional Operational Programme
to Lero—the Irish Software Research Centre (www.lero.ie).
Appendix A. Survey Instrument
This sample study was conducted via an online survey im-
plemented with SurveyMonkey. The full survey instrument is
listed below. Items prefixed with a (*) were dropped due to poor
loading onto their constructs (see Section 3 for details).
Orientation
or_1 (*) I attended an orientation program with other new hires
or_2 (*) I was made aware of the organizational rules and policies
or_3 I was assigned a buddy/mentor to help me settle in my job
or_4 I was made aware of the challenges/difficulties I may face in
my job and how I should cope with them
or_5 There were activities (like ice breakers) organized where I
could interact with my new colleagues and seniors
or_6 My company organizes team days / social gatherings to help
socialize with my colleagues
Training
tr_1 I attended a formal training program tailored for my job role
(A formal training program may involve class room based
training as a group or one-on-one training from a senior)
tr_2 I received training to understand the internal systems and
operating practices to perform my job. (Operating prac-
tices could be methodologies like Agile Programming and
Extreme Programming)
tr_3 I was specifically trained in the technology/tools that I used
for my job
tr_4 I had a point of contact/online portal that I could use if I had
any faced any difficulties regarding training
Support
su_1 If I am stuck at some task and cannot find a way through I
can ask my senior/supervisor/mentor for help
su_2 I will not feel weak or embarrassed to ask for help in the
above scenario
su_3 My supervisor provides me with ongoing constructive feed-
back about my performance
su_4 I can speak to my supervisor if any personal issues are affect-
ing my performance at work
Onboarding Success
os_1 The initial orientation program helped me feel less stressful
about joining a new workplace
14 G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
os_2 I got a good idea about the organizational culture during my
onboarding
os_3 I clearly understand the expectations and responsibilities of
my job
os_4 I am confident that I am capable of excelling in my job
os_5 I can say I am socially integrated in my workplace
Workplace Relationship Quality
rq_1 I have good professional relations with my peers
rq_2 I am good friends with some of my colleagues
rq_3 I have good relations with my seniors
Job Satisfaction
js_1 I feel that I am given a fair chance of being promoted
js_2 My work is helping my professional growth by developing
my skills and learning new technologies/tools/practices
js_3 I feel I am being given a fair compensation of the work that
I am asked to do
js_4 My performance and achievements are recognized and ap-
preciated by my senior
js_5 (*) I think that the work I am asked to do at my job is impor-
tant and meaningful
js_6 I would say that I am satisfied with my job
Turnover Intention
ti_1 I frequently think of quitting
ti_2 I will be actively looking for a new job within the next one
year
References
Aasheim, C.L., Williams, S., Butler, E.S. 2009. Knowledge and skill requirements
for IT graduates. Journal of Computer Information Systems 49 (3), 48–53.
Ajzen, I. 1991. The theory of planned behavior. Organ Behav Hum Decis Process
50 (2), 179–211.
Anderson, N.R., Cunningham-Snell, N.A., Haigh, J. 1996. Induction training as so-
cialization: Current practice and attitudes to evaluation in British organiza-
tions. International Journal of Selection and Assessment 4 (4), 169–183.
Baltes, S., Diehl, S. 2018. Towards a Theory of Software Development Expertise.
Proceedings of the 26th ACM Joint European SoftwareEngineering Conference
and Symposium on the Foundations of Software Engineering. ACM, 187–200.
Bauer, T., Erdogan, B., Bodner, T., Truxillo, D., Tucker, J. 2007. Newcomer ad-
justment during organizational socialization: a meta-analytic review of an-
tecedents, outcomes, and methods. Journal of Applied Psychology 92 (3).
Bauer, T.N. 2010. Onboarding new employees: Maximizing success. SHRM Foun-
dation’s Effective Practice Guideline Series.
Begel, A., Simon, B. 2008a. Struggles of new college graduates in their first soft-
ware development job. ACM SIGCSE Bulletin 40 (1), 226–230.
Begel, A., Simon, B. 2008b. Novice software developers, all over again. 4th Inter-
national Workshop on Computing Education Research, 3–14.
Berlin, L. 1993. Beyond Program Understanding: A Look at Programming Exper-
tise in Industry. Proceedings of the Fifth Workshop on Empirical Studies of
Programmers, 6–25.
Bolles, B.E. 2000. International student training and orientation: Current trends
in methods of programming. PhD thesis. Colorado State University.
Bourne, P.G. 1967. Some observations on the psychosocial phenomena seen in
basic training. Psychiatry 30 (2), 187–196.
Brechner, E. 2003. Things they would not teach me of in college: what Microsoft
developers learn later. Companion of the 18th annual ACM SIGPLAN confer-
ence on Object-oriented programming, systems, languages, and applications
(OOPSLA ’03). ACM, 134–136.
Britto, R., Cruzes, D.S., Smite, D., Sablis, A. 2018. Onboarding software developers
and teams in three globally distributed legacy projects: A multi-case study. J.
Softw. Evol. Proc. 30 (4), 1–17.
Brooks, F.P. 1975. The mythical man-month. Addison-Wesley.
Brown, J.S., Duguid, P. 2000. Organizational learning and communities of practice:
Toward a unified view of working, learning, and innovation. Knowledge and
communities, 99–121.
Byrne, D.J., Moore, J.L. 1997.A comparison between the recommendations of Com-
puting Curriculum 1991 and the views of software development managers in
Ireland. Comput. Educ. 28 (3), 145–154.
Cable, D.M., Gino, F., Staats, B.R. 2013. Reinventing employee onboarding. MIT
Sloan Manage. Rev. 54 (3), 23.
Canfora, G., Di Penta, M., Oliveto, R., Panichella, S. 2012. Who is going to men-
tor newcomers in open source projects? Proceedingsof the 20th International
Symposium on the Foundations of Software Engineering. ACM, 44:1–44:11.
Casado-Lumbreras, C., Colomo-Palacios, R., Soto-Acosta,P., Misra, S. 2011. Culture
dimensions in software development industry: The effects of mentoring. Sci-
entific Research and Essays 6 (11), 2403–2412.
Casalnuovo, C., Vasilescu, B., Devanbu, P., Filkov, V. 2015. Developer onboarding
in GitHub: the role of prior social links and language experience. Proceedings
of the 2015 10th joint meeting on Foundations of Software Engineering. ACM,
817–828.
Cherry, J., Arrieta, M., Brown, E., Ramaswamy, S. 2004. An Interactive Visual-
ization Tool for Understanding Complex Programs. Software Engineering Re-
search and Practice, 49–56.
Cohen, J. 1988. Statistical power analysis for the behavioral sciences. 2nd.
Dagenais, B., Ossher, H., Bellamy, R., Robillard, M., Vries, J. de. 2010. Moving into
a new software project landscape. Proceedings of the 32nd International Con-
ference on Software Engineering, 275–284.
DeMarco, T., Lister, T. 1987. Peopleware: Productive Projects and Teams. Dorset
House Publishing.
Dit, B., Revelle, M., Gethers, M., Poshyvanyk, D. 2013. Feature location in source
code: a taxonomy and survey. Journal of software: Evolution and Process 25
(1), 53–95.
Ducheneaut, N. 2005. Socialization in an open source software community: A
socio-technical analysis. Comput. Supported Coope. Work (CSCW) 14 (4), 323–
368.
Enes, P. 2005. Acquiring and Sharing Expert Knowledge. MA thesis. Norwegian
University of Science and Technology.
Fagerholm, F., Guinea, A., Münch, J., Borenstein, J. 2014a. The role of mentoring
and project characteristics for onboarding in open source software projects.
Proceedings of the 8th ACM/IEEE International Symposium on Empirical Soft-
ware Engineering and Measurement (ESEM), 55:1–55:10.
Fagerholm, F., Guinea, A.S., Borenstein, J., Münch, J. 2014b. Onboarding in open
source projects. IEEE Software 31 (6), 54–61.
Fan, J., Buckley, M.R., Litchfield, R.C. 2012. Orientation programs that may facili-
tate newcomer adjustment: A literature review and future research agenda.
Research in personnel and human resources management. Emerald Group
Publishing Limited, 87–143.
Fang, Y.,Neufeld, D. 2009. Understanding sustained participation in open source
software projects. Journal of Management Information Systems 25 (4), 9–50.
Faul, F., Erdfelder, E., Buchner, A., Lang, A.-G. 2009. Statistical power analyses us-
ing G*Power 3.1: Tests for correlation and regression analyses. Behavior Re-
search Methods 41, 1149–1160.
Feldman, D.C. 1989. tSocialization, resocialization, and training: Reframing the
research agenda. Training and Development in Organizations. Ed. by I.L. Gold-
stein. Jossey-Bass, 376–416.
Fisher, C.D. 1985. Social support and adjustment to work: A longitudinal study.
Journal of Management 11 (3), 39–53.
Forrest, C. 2018. Software had the highest job turnover rate of any indus-
try in 2017. https://www.techrepublic.com/article/software-had-the-highest-
job-turnover-rate-of-any-industry-in-2017/.
Gharehyazie,M., Posnett, D., Vasilescu, B., Filkov,V. 2015.Developer initiation and
social interactions in OSS: A case study of the Apache Software Foundation.
Empirical Software Engineering 20 (5), 1318–1353.
Graziotin, D., Fagerholm,F., Wang, X., Abrahamsson, P. 2018. What happens when
software developers are (un)happy. J. Syst. Softw. 140, 32–47.
Gundry, L.K. 1993. Fitting into technical organizations: The socialization of new-
comer engineers. IEEE Trans. Eng. Manage. 40 (4), 335–345.
Gupta, P.D., Bhattacharya, S., Sheorey, P., Coelho, P. 2018. Relationship between
onboarding experience and turnover intention: intervening role of locus of
control and self-efficacy. Industrial and Commercial Training 50 (2), 61–80.
Hair, J., Risher, J., Sarstedt, M., Ringle, C. 2019. When to use and how to report the
results of PLS-SEM. European Business Review 31 (1), 2–24.
Hair, J.F., Ringle, C.M., Sarstedt, M. 2011. PLS-SEM: Indeed a silver bullet. Journal
of Marketing theory and Practice 19 (2), 139–152.
Hair, J.F., Hult, G.T.M., Ringle, C., Sarstedt, M. 2016. A primer on partial least
squares structural equation modeling (PLS-SEM). Sage.
Hall, T., Beecham, S., Verner, J., Wilson, D. 2008. The Impact of Staff Turnover on
Software Projects: The Importance of Understanding What Makes Software
Practitioners Tick. Proceedings of ACM SIGMIS CPR. ACM, 30–39.
Henseler, J., Ringle, C.M., Sinkovics,R.R. 20 09. The use of partial least squarespath
modeling in international marketing. New challenges tointernational market-
ing. Emerald Group Publishing Limited, 277–319.
Henseler, J., Ringle, C.M., Sarstedt, M. 2015. A new criterion for assessing discrim-
inant validity in variance-based structural equation modeling. J. Acad. Mark.
Sci. 43 (1), 115–135.
Hertzum, M., Pejtersen, A.M. 2000. The information-seeking practices of engi-
neers: searching for documents as well as for people. Information Processing
& Management 36 (5), 761–778.
Hilton, M., Begel, A. 2018. A Study of Organizational Dynamics of Software Teams.
Proceedings of the ACM/IEEE 40th International Conference on Software Engi-
neering: Software Engineering in Practice. ACM, 191–200.
Holtom, B.C., Mitchell, T.R., Lee, T.W., Eberly, M.B. 2008. Turnover and Retention
Research: A Glance at the Past, a Closer Review of the Present, and a Venture
into the Future. The Academy of Management Annals 2 (1), 231–274.
Hom, P.W., Lee, T.W., Shaw, J.D., Hausknecht, J.P. 2017. One hundred years of em-
ployee turnover theory and research. Journal of Applied Psychology 102 (3),
530–545.
Johnson, M., Senges, M. 2010. Learning to be a programmer in a complex organi-
zation: A case study on practice-based learning during the onboarding process
at Google. Journal of Workplace Learning 22 (3), 180–194.
G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442 15
Klein, H.J., Fan, J., Preacher, K.J. 2006. The effects of early socialization experiences
on content mastery and outcomes: A mediational approach. Journal of Voca-
tional Behavior 68 (1), 96–115.
Klein, H.J., Weaver, N.A. 2000. The effectiveness of an organizational-level orien-
tation training program in the socialization of newhires. Personnel Psychology
53 (1), 47–66.
Korte, R., Lin, S. 2013. Getting on board: Organizational socialization and the con-
tribution of social capital. Human Relations 66 (3), 407–428.
LaToza, T.D., Venolia, G., DeLine, R. 2006. Maintaining mental models: a study of
developer work habits. 28th International Conference on Software Engineer-
ing, 492–501.
Lave, J., Wenger, E. 1991. Situated learning: Legitimate Peripheral Participation.
Cambridge University Press.
Lavigna, B. 2009. Getting onboard: Integrating and engaging new employees.
Government Finance Review 25 (3), 65–70.
Lee, S., Fang, X. 2008. Perception Gaps about Skills Requirement for Entry-Level
IS Professionals between Recruiters and Students: An Exploratory Study. Inf.
Resour. Manage. J. 21 (3), 39–63.
Lee, T.W., Mitchell, T.R. 1994. An Alternative Approach: The Unfolding Model of
Voluntary EmployeeTurnover. Academyof Management Review 19 (1), 51–89.
Legier, J., Woodward, B., Martin, N.L. 2013. Reassessing the skills required of grad-
uates of an information systems program: An updated analysis. Information
Systems Education Journal 11 (3).
Lenberg, P., Tengberg, L.G.W., Feldt, R. 2017. An Initial Analysis of Software Engi-
neers’ Attitudes Towards Organizational Change. Empirical Software Engineer-
ing 22 (4), 2179–2205.
Lethbridge, T.C., Singer, J., Forward, A. 2003. How software engineers use docu-
mentation: The state of the practice. IEEE software 6, 35–39.
Lethbridge, T.C. 1998. A Survey of the Relevance of Computer Science and Soft-
ware Engineering Education. Proceedings of the 11th Conference on Software
Engineering Education and Training (CSEET ’98). IEEE Computer Society, 56–
66.
Lethbridge, T.C. 2000. Priorities for the education and training of software engi-
neers. The Journal of Systems and Software 53 (1), 53–71.
Littman, D.C., Pinto, J., Letovsky, S., Soloway,E. 1987. Mental models and software
maintenance. Journal of Systems and Software 7 (4), 341–355.
Louis, M.R., Posner, B.Z., Powell, G.N. 1983. The availability and helpfulness of so-
cialization practices. Personnel Psychology 36 (4), 857–866.
Marcoulides, G.A., Saunders, C. 2006. Editor’s Comments: PLS: A Silver Bullet?
MIS Quarterly 30 (2), iii–ix.
McGuire, E.G., Randall, K.A. 1998. Process improvement competencies for IS pro-
fessionals: a survey of perceived needs. Proceedings of the 1998 ACM SIGCPR
conference on Computer personnel research (SIGCPR ’98). Ed. by R. Agarwal.
ACM, 1–8.
McMurtrey, M.E., Downey, J.P., Zeltmann, S.M., Friedman, W.H. 2008. Critical skill
sets of entry-level IT professionals: An empirical examination of perceptions
from field personnel. J. Inf. Technol. Educ. 7, 101–120.
Metcalf, A.Y., Stoller, J., Habermann, M., Fry, T. 2015. Respiratory therapist job per-
ceptions: the impact of protocol use. Respiratory care 60 (11).
Miller, J. 2018. Why Do Software Engineers Change Jobs So Frequently?
https://www.forbes.com/sites/quora/2018/02/06/why-do-software-
engineers-change-jobs-so-frequently/.
Nelson, D.L., Quick, J.C. 1991. Social support and new- comer adjustment in orga-
nizations: Attachment theory at work? J Org Behav 12 (6), 543–554.
Nitzl, C., Roldán, J., Cepeda, C. 2016. Mediation analysis in partial least squares
path modeling: Helping researchers discuss moresophisticated models. Indus-
trial Management and Data Systems 119 (9), 1849–1864.
O’Brien, M. 2007. Evolving a Model of the Information-Seeking Behaviour of In-
dustrial Programmers. PhD thesis. University of Limerick.
Panichella, S. 2015. Supporting newcomers in software development projects.
IEEE International Conference on Software Maintenance and Evolution (IC-
SME), 586–589.
Papanikolaou, A., Karakoidas, V., Vlachos, V., Venieris, A., Ilioudis, C., Zouganelis,
G. 2011. A Hacker’s Perspective on Educating Future Security Experts. 5th Pan-
hellenic Conference on Informatics with international participation (PCI 2011),
68–72.
Parolia, N., Chen, J.V., Jiang, J.J., Klein, G. 2015. Conflict resolution effectiveness
on the implementation efficiency and achievement of business objectives in
IT programs: A study of IT vendors. Inform Softw Technol 66, 30–39.
Parolia, N., Jiang, J.J., Klein, G. 2013. The presence and development of competency
in IT programs. J Syst Softw 86, 3140–3150.
Pham, R., Kiesling, S., Singler, L., Schneider, K. 2017. Onboarding inexperienced
developers: struggles and perceptions regarding automated testing. Software
Quality Journal 25 (4), 1239–1268.
Pham, R., Stoliar, Y., Schneider, K. 2015. Automatically recommending test code
examples to inexperienced developers. 10th Joint Meeting on Foundations of
Software Engineering, 890–893.
Qureshi, I., Fang, Y. 2011.Socialization in open source software projects: A growth
mixture modeling approach. Org Res Meth 14 (1), 208–238.
Radermacher, A., Walia, G. 2013. Gaps between industry expectations and the
abilities of graduates. Proceeding of the 44th ACM technical symposium on
Computer science education (SIGCSE ’13). ACM, 525–530.
Rastogi, A., Thummalapenta, S., Zimmermann, T., Nagappan, N., Czerwonka, J.
2015. Ramp-up Journey of New Hires Tug of War of Aids and Impediments.
Proceedings of the ACM/IEEE International Symposium on Empirical Software
Engineering and Measurement (ESEM), 96–105.
Rawlins, W. 1992. Friendship matters: Communication, dialectics, and the life
course. Aldine de Gruyter.
Russo, D., Stol, K.-J. 2019. Soft theory: a pragmatic alternative to conduct quanti-
tative empirical studies. Proceedings of the Joint 7th International Workshop
on Conducting Empirical Studies in Industry and 6th International Workshop
on Software Engineering Research and Industrial Practice. IEEE Press, 30–33.
Saks, A.M. 1995. Longitudinal field investigation of the moderating and mediat-
ing effects of self-efficacy on the relationship between training and newcomer
adjustment. J. Appl. Psychol. 80 (2).
Sarma, A., Gerosa, M., Steinmacher, I., Leano, R. 2016. Training the Future Work-
force through TaskCuration in an OSS Ecosystem. Proceedings of the 201624th
ACM SIGSOFT International Symposium on Foundations of Software Engineer-
ing, 932–935.
Sias, P., Cahill, D. 2013. From coworkers to friends: The development of peer
friendships in the workplace. Western Journal of Communication 62 (3), 273–
299.
Sias, P.M. 2005. Workplace Relationship Quality and Employee Information Ex-
periences. Communication Studies 56 (4), 375–395.
Sillito, J., Murphy,G.C., Volder, K.D. 2008. Asking and answering questions during
a programming change task. IEEE Trans Softw Eng 34 (4), 434–451.
Sim, S.E., Holt, R. 1998. The Ramp-Up Problem in Software Projects: A Case Study
of How Software Immigrants Naturalize. Proceedings of the Twentieth Inter-
national Conference on Software Engineering, 361–370.
Simmons, C.B., Simmons, L.L. 2010. Gaps in the computer science curriculum: an
exploratory study of industry professionals. Journal of Computing Sciences in
Colleges 25 (5), 60–65.
Simon, D., Jackson, K. 2013. A closer look at information systems graduate prepa-
ration and job needs: Implications for higher education curriculum enhance-
ments. World Journal of Education 3 (3), 52–62.
Snell, A. 2006. Researching onboarding best practice: using research to connect
onboarding processes with employee satisfaction. Strategic HR Review 5 (6),
32–35.
Spector, P.E. 1985. Measurement of human service staff satisfaction: Develop-
ment of the Job Satisfaction Survey. American journal of community psychol-
ogy 13 (6), 693–713.
Steinmacher, I., Silva, M.A.G., Gerosa, M.A. 2014. Barriers faced by newcomers to
open source projects: a systematic review. IFIP International Conference on
Open Source Systems, 153–163.
Stol, K.-J., Fitzgerald, B. 2018. The ABC of Software Engineering Research. ACM
Trans Softw Engineer Methodol 27 (3).
Surakka, S. 2007.What subjects and skills are important for software developers?
Communications of the ACM 50 (1), 73–78.
Tang, H.-L., Lee, S., Koh, S. 2001. Educational gaps as perceived by IS educators: A
survey of knowledge and skill requirements. Journal of Computer Information
Systems 41 (2), 76–84.
Tesch, D., Braun, G., Crable, E. 2008. An examination of employers’ perceptions
and expectations of IS entry-level personal and interpersonal skills. Informa-
tion Systems Education Journal 6 (1), 3–16.
Tett, R.P., Meyer, J.P. 1993. Job satisfaction, organizational commitment, turnover
intention, and turnover: path analyses based on meta-analytic findings. Per-
sonnel psychology 46 (2), 259–293.
Trent, R.H. 1988. Perspectives on the academic preparation of MIS professionals.
Proceedings of the ACM SIGCPR conference on Management of information
systems personnel (SIGCPR ’88). ACM, 119–119.
Van Maanen, J. 1978. People processing: Major strategies of organizational so-
cialization and their consequences. New directions in human resource man-
agement. Ed. by J. Paap. Prentice Hall, 19–36.
Van Maanen, J.E., Schein, E.H. 1979. Toward a theory of organizational socializa-
tion. Research in organizational behavior. Ed. by B. Staw. Vol. 1. JAI Press, 209–
264.
Viana, D., Conte, T., Souza, C.R.B. de. 2014. Knowledge Transfer between Senior
and Novice Software Engineers: A Qualitative Analysis. Proceedings of the
26th International Conference on Software Engineering and Knowledge Engi-
neering, 235–240.
Vijayasarathy, L., Turk, D. 2012. Drivers of agile software development use: Di-
alectic interplay between benefits and hindrances. Information and Software
Technology 54, 137–148.
Von Mayrhauser, A., Vans, A.M., Howe, A.E. 1997. Program understanding be-
haviour during enhancement of large-scale software. Journal of Software
Maintenance: Research and Practice 9 (5), 299–327.
Wanous, J.P. 1992. Organizational entry:Recruitment, selection, orientation, and
socialization of newcomers. Prentice Hall.
Wanous, J.P., Reichers, A.E. 2000. New employee orientation programs. Human
resource management review 10 (4), 435–451.
Watkins, M. 2013. The first 90 days, updated and expanded: provenstrategies for
getting up to speed faster and smarter. Harvard Business Review Press.
Wesson, M.J., Gogus, C.I. 2005. Shaking hands with a computer: an examination
of two methods of organizational newcomer orientation. Journal of Applied
Psychology 90 (5).
Yates, R., Power, N., Buckley, J. 2020. Characterizing the transfer of program com-
prehension in onboarding: an information-push perspective. Empirical Soft-
ware Engineering 25, 940–995.
Zhao, X., Lynch Jr.,J.G., Chen, Q. 2010. Reconsidering Baron and Kenny: Myths and
Truths about Mediation Analysis. Journal of Consumer Research 37, 197–206.
16 G.G. Sharma and K.-J. Stol / The Journal of Systems and Software 159 (2020) 110442
Gaurav G. Sharma is a DevOps Engineer in Security Intelligence with IBM in Ire-
land. He was the co-founder of an e-commerce start-up called Pocket My Need
in his final year of engineering. He has a keen interest in project management,
automation, and DevOps methodologies. He received a M.Sc. in Computer
Science from University College Cork, and a B.Eng. in Information Technology
from the University of Mumbai. Contact him at gaurav.gyan.sharma@ibm.com.
Klaas-Jan Stol is a lecturer with the School of Computer Science and Informa-
tion Technology at University College Cork. He is a Science Foundation Ireland
(SFI) Principal Investigator, and an SFI Funded Investigator with Lero, the Irish
Software Research Centre. His research focuses on contemporary software en-
gineering strategies and research methodology. Stol received a Ph.D. from the
University of Limerick and a M.Sc. from the University of Groningen. Contact
him at klaas-jan.stol@lero.ie.
