Autonomy and Turnover: A Survey Applied to Distributed Software

Teams

Luis Amorim

1 a

, Ivaldir de Farias J

unior

2 b

and Marcelo Marinho

1 c

Departamento de Computac¸

ao (DC), Universidade Federal Rural de Pernambuco (UFRPE), 52071-000, Recife, PE, Brazil

Departamento de Computac¸

ao, Universidade de Pernambuco (UPE), Garanhuns, PE, Brazil

Keywords:

Autonomy, Turnover, Distributed Software Development.

Abstract:

Distributed teams have gained prominence in software companies. However, studies indicate that Distributed

Software Development (DSD) companies often face challenges related to high developer turnover. Conversely,

other research suggests that autonomy and its associated factors can mitigate or prevent such turnover. This

study investigates the relationship between autonomy and turnover within DSD teams. To accomplish this,

we conducted a survey based on previous Systematic Literature Review (SLR) research involving 181 soft-

ware engineers worldwide. Our ﬁndings shed light on the key autonomy factors that impact turnover in DSD

projects, including recognition, communication, collaboration, trust, and task balance. By offering a compre-

hensive understanding of these autonomy factors, our study provides software companies and organizations

with valuable insights for addressing the issue of turnover in DSD projects.

1 INTRODUCTION

Software development has become a pivotal driver

of innovation and business growth in today’s global-

ized economies and interconnected markets. Organi-

zations increasingly adopt a Distributed Software De-

velopment (DSD) model to harness diverse talent, op-

timize costs, and achieve faster time-to-market (Mar-

inho et al., 2018). DSD allows companies to leverage

expertise from different geographical locations, fos-

tering collaboration among distributed teams (Mar-

inho et al., 2019). However, managing DSD teams

poses unique challenges, including the dynamic inter-

play between autonomy, turnover, and project success

(Bass et al., 2018).

Autonomy could be deﬁned from different per-

spectives, such as individual autonomy, internal and

external autonomy, where external autonomy is de-

ﬁned as the inﬂuence of management and other indi-

viduals outside the team on the team’s activities. In-

ternal autonomy refers to the degree to which all team

members jointly share decision authority, while indi-

vidual autonomy refers to the freedom and discretion

an individual has in carrying out assigned tasks (Moe

https://orcid.org/0000-0001-8970-4204

https://orcid.org/0000-0001-9860-8206

https://orcid.org/0000-0001-9575-8161

et al., 2021).

According to Fitzgerald et al. (Fitzgerald et al.,

2017), high levels of team autonomy in DSD can

foster creativity, initiative, and ownership, enabling

teams to adapt quickly to local challenges and op-

portunities. On the other hand, the issue of turnover

within DSD teams, encompassing the addition or

departure of team members, presents a substantial

hurdle. The dispersed structure of such teams can

magnify the impact of turnover due to communica-

tion and coordination complexities. Employee at-

trition can potentially be highly disruptive, resulting

in the erosion of knowledge, diminished team unity,

and heightened project vulnerability (Massoni et al.,

2019).

In this study, we aim to investigate the relationship

between autonomy and turnover within DSD teams,

considering the perceived impacts of this relationship.

To achieve this, we have designed a survey based on

a systematic literature review (Chaves et al., 2022)

that we have previously conducted. The survey aims

to provide a broader understanding of autonomy-

turnover dynamics by addressing the following re-

search questions: (RQ01) How do software engineers

within DSD teams perceive autonomy? (RQ02) What

do software engineers perceive as these teams’ most

signiﬁcant turnover motivators? and (RQ03) How

does autonomy impact turnover within DSD teams?

134

Amorim, L., Farias Júnior, I. and Marinho, M.

Autonomy and Turnover: A Survey Applied to Distributed Software Teams.

DOI: 10.5220/0012162300003598

In Proceedings of the 15th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2023) - Volume 3: KMIS, pages 134-141

ISBN: 978-989-758-671-2; ISSN: 2184-3228

By exploring the interplay between autonomy and

turnover within DSD teams, our study aims to con-

tribute to a better understanding of these complex dy-

namics. The ﬁndings have practical implications for

organizations, offering insights into the perceived au-

tonomy factors and turnover motivators that inﬂuence

software engineers in distributed settings. Ultimately,

this research aims to provide valuable knowledge that

can assist software companies in addressing the chal-

lenges associated with turnover in DSD teams.

Our study is organized as follows: in Section 2,

we introduce the background. Section 3 describes the

method to design this survey. Section 4 displays the

gathered results from participants. Then, Section 5

presents discussions about results and literature. Sec-

tion 6 presents the limitations of our study, Section 7

highlights our conclusion and possible further inves-

tigations. Yet, in Section 7 we also have our Survey

protocol available.

2 BACKGROUND

2.1 Autonomy

Autonomy, a pivotal aspect of work design and moti-

vation, has been scrutinized in research (Noll et al.,

2017). Initially, autonomy was synonymous with

the extent of freedom and independence granted to

team members during project execution (Hackman

and Oldham, 1976). Subsequent investigations (Bass

et al., 2018; Marinho et al., 2021) expounded upon

this concept, highlighting its connection to traits such

as decision-making authority, freedom, and the ability

to shape work routines and methodologies.

Autonomy’s signiﬁcance extends to DSD, where

professionals contend with unique challenges (Mar-

inho et al., 2018). In this context, autonomy is both

a motivator and a potential source of dissatisfaction.

Its absence can amplify the need for autonomy and

increase dissatisfaction, possibly leading to turnover

(Bass et al., 2018). For software engineers operat-

ing within distributed models, autonomy emerges as a

critical motivational factor, particularly given the de-

mands of antisocial work hours and travel (Noll et al.,

2017).

The relationship between autonomy and turnover

intention within DSD teams has been a subject of

inquiry (Dysvik and Kuvaas, 2013). Furthermore,

the lack of autonomy has been identiﬁed as a poten-

tial contributor to decreased job satisfaction among

software engineers, which may ultimately lead to

turnover (Noll et al., 2017).

2.2 Turnover

The disruptive effects of team member turnover on

software development processes are widely acknowl-

edged. Although some experts acknowledge turnover

as a natural organizational process, its excessive oc-

currence affects process efﬁciency adversely (Bass

et al., 2018). This recognition underscores the criti-

cality of managing turnover within software develop-

ment teams.

Turnover can be categorized as either external or

internal. In external turnover, team members exit

the organization entirely, while internal turnover in-

volves members remaining within the company but

altering their roles (Chatzipetrou et al., 2018). Vol-

untary turnover occurs when employees opt to leave

their positions and the company, whereas involuntary

turnover transpires when organizations terminate em-

ployee contracts (Chatzipetrou et al., 2018).

Organizations often calculate turnover rates by di-

viding the number of departures within a speciﬁc

timeframe by the total workforce during that period to

gauge turnover. However, more comprehensive and

nuanced data is essential for effective strategy for-

mulation. This data, highlighting the underlying rea-

sons for turnover, is a crucial asset for creating strate-

gies that effectively manage and mitigate the conse-

quences of turnover (Chatzipetrou et al., 2018). In

the context of DSD teams, turnover risk is accentu-

ated due to physical distance, further magnifying its

importance as one of the foremost risks in DSD ven-

tures (Ebert et al., 2016).

2.3 Distributed Software Development

(DSD)

A distributed project is a collaborative effort where

individuals from different locations work together on

a single project over an extended period. This type

of software project, which involves human resources

spread across different distances, regions, or even

countries, is known as Distributed Software Develop-

ment (DSD) (Marinho et al., 2019).

DSD can be classiﬁed based on two factors. First,

the distance among work teams is categorized as On-

shore (within the same country) or Offshore (in dif-

ferent countries). Second, the control relationship the

central organization has over the remote teams distin-

guishes it as either Outsourcing (involving the hiring

of a third-party company) or Insourcing (establishing

a remote unit within the company) (Bass et al., 2018).

Research suggests that offshore outsourcing ser-

vice providers can mitigate staff turnover by pri-

oritizing work-life balance and implementing more

Autonomy and Turnover: A Survey Applied to Distributed Software Teams

135

family-friendly employment policies, such as grant-

ing increased autonomy. Furthermore, these ser-

vice providers can effectively incentivize innovation

and structure contracts to enhance software product

ownership, ultimately improving staff retention (Bass

et al., 2018).

3 METHOD

3.1 Survey Design

The basic idea of survey methodology is to collect

information from a group of people by sampling in-

dividuals from a large population. Examples of sur-

veys are found in daily life in several situations, such

as election polls, market surveys, etc. There is a

large amount of literature on the general methodology

(Linaker et al., 2015).

Therefore, this survey based its assumptions and

questions on results gathered by previously executed

SLR (Chaves et al., 2022) and aimed to assess the per-

ception of turnover, autonomy, and their relationship

among other related factors within DSD teams such

as payment, growth opportunity, work-life balance,

collaborative environment, good communication, su-

pervisor support, stress, motivation, good and active

leadership, freedom for decision making, connection

with co-workers, the balance of tasks, satisfaction,

clear career orientation.

This survey was divided into four subsections. Ini-

tially, three fundamental questions were displayed,

aiming at a certain level of participant evaluation and

their distributed teams’ autonomy. They all were

stated using the 5-point Likert scale where one means

’Strongly disagree’ and ﬁve means ’Totally agree’.

In the second section, there was a goal to iden-

tify turnover-related results. Therefore, for this sec-

tion, it was possible to have questions more focused

on turnover intentions and what could be improved

inside DSD teams and questions focused on impacts

generated by turnover in DSD teams. Furthermore,

this section was designed with open questions and

classiﬁcation ones.

The third section displayed open questions and af-

ﬁrmative statements using the 5-point Likert scale,

such as in the ﬁrst section. Participants evaluated

those questions and statements for further analy-

sis to check the relationship between autonomy and

turnover/turnover intentions.

Ultimately, the study’s demography perspective

was addressed with questions intended to collect par-

ticipant’s gender, age, work experience, education,

role, team distribution, and company size.

3.2 Setting

This study aimed to collect data on turnover and au-

tonomy within DSD teams. A web survey was con-

ducted using the Google Forms tool to achieve this. A

pilot session involving three software engineers was

conducted to ensure the survey’s quality, readabil-

ity, and validity. Their feedback was invaluable in

addressing any shortcomings and reﬁning statements

and questions.

In addition to the pilot session, multiple web plat-

forms distributed the survey to participants, including

LinkedIn, Twitter, WhatsApp, email, and Instagram.

Despite the time constraint – the survey was con-

ducted from August 20th to September 30th – a sub-

stantial response was achieved. 181 participants com-

pleted the survey, excluding the three from the pilot

phase. This strong response underscores the survey’s

effectiveness even within a limited timeframe. The

paper reports a qualitative study on the relationship

between autonomy and turnover within DSD teams.

Therefore, no statistical power is required.

3.3 Participants

The 181 participants previously mentioned were

mostly software developers, testers, scrum masters,

project managers and leaders, technology leaders, se-

curity analysts, UX/UI designers, software architects,

data scientists, etc. They all had at least one per-

son from their teams working in a distributed envi-

ronment.

3.4 Analysis

Yet, for the data analysis procedure, the author chose

a qualitative approach where all collected data was ex-

tensively analyzed, generating perceptions stored in a

spreadsheet, which was later on translated into charts,

tables, and quotes, constructing parallelisms by en-

gaging in triangulation to support arguments and con-

clusions for this study (Linaker et al., 2015).

4 RESULTS

4.1 Study Population

The survey yielded complete answers from software

engineers involved in the software development life

cycle regarding autonomy, turnover, and their rela-

tionship inside distributed environments, with 85 par-

ticipants (47%) working remotely inside Brazil, 76

KMIS 2023 - 15th International Conference on Knowledge Management and Information Systems

136

participants (42%) working with other software en-

gineers globally located, 13 participants (7%) who

worked with team members from the same continent

and 7 participants (3%) working in locally distributed

teams within the same city or state.

Most of the identiﬁed participants were men

(69.1%), some were women (28.7%) and completing

the total number of participants, there were 1.1% non-

binary and 1.1% of not identiﬁed gender.

87.3% of the participants were found to work

in a large company, 3.9% worked in a small com-

pany, 6.1% worked in a medium company and 2.8%

worked in a micro company. Furthermore, a range

of experience in the software development area were

stated amongst them ranging from less than 1 year (6),

through 1-5 years (92), 6-10 years (36), 10-15 years

(22) to more than 15 years (14). Yet, 54.1% of the

participants were software developers, 22.7% were

testers, 5.5% were software architects, 4.4% were re-

quirements engineers and the rest of them were split

among management/leadership roles, UX/UI, Scrum

Masters, trainees and other roles.

4.2 RQ01. How Do Software Engineers

Within DSD Teams Perceive

Autonomy?

It was possible to notice a pretty well-balanced col-

lected data where 50 (27.6%) participants stated that

they did not know whether their work was being

strongly monitored or not, while 63 (34%) declared

there is at least some strong monitoring of their work

and 68 (37%) of the participants informed there is

few monitoring within their work in software devel-

opment cycles.

Regarding teams having a shared purpose, clear

goal, necessary skills, and mutual trust among peers,

160 (88.3%) participants agreed at least partially to it.

In contrast, only 12 participants (6.6%) denied it, and

the other nine stated they did not know whether this

happened within their teams.

Yet, results show that 156 (86.1%) participants

agree they have at least some healthy support and

freedom to make decisions provided by managers and

leaders. On the other hand, 12 participants (6.6%) de-

nied it, and 13 stated not knowing about any support

or freedom from leadership members.

4.3 RQ02. Which Turnover Motivators

Are the Most Perceived for Software

Engineers Within DSD Teams?

To assess the causes and mitigation factors of turnover

within distributed teams, we presented a section to the

participants regarding this matter with questions and

statements.

According to results, 85.5% of all 181 participants

stated they have already decided to leave a project or

job, while 14.5% of them have never been through

such an experience.

Participants could also select the most critical fac-

tors in deciding whether to leave a project or com-

pany. The outcomes revealed a ranking of perceived

factors such as payment, stress, lack of motivation,

the mismatch between expectation and reality, lack

of motivation, lack of supervisor support, long work-

ing hours, lack of satisfaction, poor communication

among peers, lack of freedom for decision making

and lack of connection with co-workers as can be seen

on Table 1.

Table 1: Most important factors to leave a project or com-

pany according to participants (n=181).

Option # %

Payment 145 80

Stress 130 72

Lack of motivation 112 62

Mismatch between expectations and reality 106 59

Lack of collaboration among the team 101 56

Lack of supervisor support 97 54

Long working hours 92 51

Lack of satisfaction 88 49

Lack of connection with co-workers 73 40

Lack of freedom for decision making 73 40

On the other hand, participants could select the

factors they considered the most important in decid-

ing whether to stay in a project or company. Based on

their answers, another ranking was lightened with per-

ceived factors such as payment, growth opportunity,

work-life balance, collaborative environment, good

communication, supervisor support, motivation, good

and active leadership, freedom for decision making,

connection with co-workers, the balance of tasks, sat-

isfaction, clear career orientation, all team members

feeling involved, organizational commitment, work-

place innovation, and employment policies as can be

seen on Table 2.

Yet, 90.4% of them declared they had already been

to a distributed project or company where a software

engineer left, while 9.5% stated this never happened

before. Also, all participants stated the most com-

mon outcomes after a person leaves a project or com-

Autonomy and Turnover: A Survey Applied to Distributed Software Teams

137

Table 2: Most important factors to staying in a project or

company according to participants (n=181).

Option # %

Payment 158 87

Growth opportunity 153 85

Work-life balance 143 79

Collaborative environment 136 75

Good communication 119 66

Supervisor support 115 64

Motivation 110 61

Good and active leadership 107 59

Freedom for decision making 102 56

Connection with co-workers 101 56

Balance of tasks along development cycles 93 51

Satisfaction 90 50

Clear career orientation 79 44

All team members feeling involved 73 40

Organizational commitment 67 37

Workplace innovation 50 28

Employment policies 24 13

Table 3: Most important outcomes after a software engi-

neer leaves a project or company according to participants

(n=181).

Option # %

Loss of knowledge and experience 154 85

Lower levels of productivity 78 43

Software quality 57 31

Lack of commitment and mutual trust among peers 41 23

Economic loss to companies 28 15

Better conditions and performance for those who remain 24 13

Fresh and innovative ideas 18 10

Unsuccess of software project 17 9

Project success 8 4

pany the participant is still a part of. Those outcomes

were also brought up in a ranking and contained items

such as lower levels of productivity, project success,

loss of knowledge and experience, software quality,

better conditions and performance for those who re-

main, economic loss to companies, lack of commit-

ment, and mutual trust among peers, fresh and inno-

vative ideas, software project unsuccessful as can be

seen in Table 3.

4.4 RQ03. How Does Autonomy Impact

Turnover Inside DSD Teams?

Our study delved into the intricate relationship be-

tween autonomy and turnover as perceived by our par-

ticipants. Initially, we questioned whether a higher

level of autonomy within distributed projects was a

key factor inﬂuencing one’s willingness to stay or

leave, and the results showed that 69.8% of partici-

pants agreed, while 30.2% did not. To gain deeper

insights, we explored how participants in different

career stages perceived autonomy. We invited them

to share their real experiences, focusing on moments

when they were granted autonomy early in their ca-

reers or as more experienced professionals.

During early career stages, participants reported

several positive outcomes associated with autonomy.

Trust was a signiﬁcant theme, as many felt trusted by

their team members, which fostered their growth and

development. Recognition also ﬂourished, with par-

ticipants taking on more responsibilities and feeling

acknowledged for their skills. Autonomy was seen as

a catalyst for improved communication and collabo-

ration, even when individuals had limited knowledge.

Support from more experienced engineers played a

pivotal role, acting as mentors and guiding less expe-

rienced team members. Furthermore, autonomy was

motivating, making participants feel valued and in-

spired.

However, it was not all positive. Some partici-

pants shared initial feelings of anxiety, fear, and inse-

curity when granted autonomy. These negative emo-

tions often dissipated over time as they gained expe-

rience and received support from their colleagues.

In contrast, participants in more advanced stages

of their careers emphasized the importance of auton-

omy in decision-making. They valued the freedom

to choose the best approaches for their tasks. With

autonomy came increased accountability, leading to

professional growth and satisfaction, although some

experienced initial nervousness. Autonomy consis-

tently correlated with professional growth, whether

during task execution or afterward. Even though au-

tonomy was highly regarded, participants acknowl-

edged the ongoing need for communication and col-

laboration within their teams. Trust played a dual

role—trust from the team to grant autonomy and trust

among more experienced engineers to promote auton-

omy within the team.

Moreover, we presented seven statements to gauge

participants’ perceptions of autonomy and its link to

turnover intentions using a Likert scale. The ﬁndings

revealed that most participants believed that having

the freedom to work as they preferred heightened their

motivation, satisfaction, and willingness to remain in

a project or company. Surprisingly, a lack of compe-

tence was not signiﬁcantly associated with stress and

turnover intentions when autonomy was present. In-

stead, participants viewed challenges as opportunities

for personal and professional growth. Autonomy also

positively inﬂuenced task control, quality, and team

connections. Participants tended to choose challeng-

ing tasks when given the option, leading to increased

delight in their work. Autonomy fostered individual

trust and commitment to team goals, boosting engage-

ment and potentially reducing turnover.

KMIS 2023 - 15th International Conference on Knowledge Management and Information Systems

138

Interestingly, many participants preferred working

in a quiet environment when handling complex tasks

autonomously, as it reduced exhaustion. This pref-

erence for solitude was not necessarily a drawback,

especially in distributed work environments. Lastly,

participants generally favored tasks related to new

products or innovative ideas over support or legacy

items, signaling higher engagement and the potential

for reduced turnover rates within software teams.

5 DISCUSSION

Autonomy and turnover have been widely discussed

in previous researches (Lin et al., 2017; Noll et al.,

2017; Chaves et al., 2022; Marinho et al., 2021) sepa-

rately and usually linked to some factors such as mo-

tivation, satisfaction, happiness, communication, col-

laboration, and many others. In our work, we ad-

dressed the perception about the relationship between

them.

Most software engineers believe the organization

is becoming too intrusive of their private space, and

many would want to disassociate themselves from

employment entirely and set up their businesses in-

stead, just to be able to exercise their freedom (Uzoka

et al., 2011). Freedom for decision-making has been

cited in our results as one of the most important fac-

tors linked to autonomy in distributed teams, espe-

cially for those more experienced software engineers,

but there was no indication of disassociation with

companies. Instead, there were many indications that

this freedom is strongly linked to trust and account-

ability. This mutual trust among peers and sense of

accountability enables them to share more activities,

consequently fostering them to share information so

everyone has the knowledge to inﬂuence decisions

(Robert Jr and You, 2018; Lundene and Mohagheghi,

2018).

This connection among peers has already been

identiﬁed as both good and bad regarding turnover in-

tentions as it may retain software engineers working

together for a longer period (Bass et al., 2018) but on

the other hand, it may become an inﬂuence for higher

turnover as the career moves of co-workers strongly

inﬂuence individuals in attempts to deﬁne security,

stability, and career success (Moe et al., 2021).

Even though autonomy is not a problem for those

working remotely as it can work independently as a

pre-requisite (Noll et al., 2017), it’s also found to have

a strong link to communication inside distributed soft-

ware teams (Noll et al., 2017). Yet, a lack of ade-

quate communication inside a company is commonly

related to low levels of commitment and high turnover

levels (Noll et al., 2017). In our study, we could iden-

tify the need for communication stated by those soft-

ware engineers who have been given autonomy at the

workplace in different career stages. Also, good com-

munication is an important factor in staying in a dis-

tributed software team. Therefore, it supports the vi-

sion brought by Bass et al. (Bass et al., 2018) where

poor communication is related to a lot of tension and

pressure, which may lead to stress; this stress can

have a considerable impact on organizational commit-

ment and turnover levels (Yener et al., 2020; Hynni-

nen et al., 2010).

Collaboration reduces workplace tension and

stress levels among the team (Bass et al., 2018). A

collaborative environment is an important factor for

participants’ decision to stay in a project. When col-

laboration is lacking, it can lead to workload imbal-

ances and potentially increase project turnover (Bao

et al., 2017). Experienced participants in our study

mentioned that autonomy can result in task overload

due to increased accountability. Lack of collabora-

tion can also impact early-career software engineers

who rely on support from senior colleagues for skill

development (Noll et al., 2017). Additionally, lack

of collaboration can affect work estimation, manage-

ment, and team environment, inﬂuencing retention

(Bass et al., 2018). Our results indicate that lack of

collaboration signiﬁcantly inﬂuences participants’ in-

clination to leave a project or company.

However, recognition was a positive outcome

mentioned by participants who have experienced au-

tonomy as they feel valued and involved. Autonomy,

pay level, promotional chances, and social support are

suggested to be positively related to organizational

commitment and thus reduce the likelihood of volun-

tary turnover (Hynninen et al., 2010) while according

to Uzoka et al. (Uzoka et al., 2011), good payment

and promotion can be considered a kind of recogni-

tion for the services provided, and they inﬂuence the

intention to leave. Our work indicates that it is not

only payment is related to turnover intentions but also

to turnover retention.

This perception is quite well aligned with the

appreciation for challenges stated in (Remus et al.,

2016), and it indicates the positive impact on job

and career satisfaction after opportunities for career

development, promotions, and training opportunities

(Uzoka et al., 2011). Therefore, even though pay-

ment is identiﬁed as a factor linked to turnover, au-

tonomy can be considered a factor that leads to pro-

fessional growth opportunities, motivation, and satis-

faction, which can be as important as payment.

Yet, other factors linked to autonomy and turnover

are present, such as growth opportunities and motiva-

Autonomy and Turnover: A Survey Applied to Distributed Software Teams

139

tion. Some participants have stated that the autonomy

received by them acted as a motivator during working

days and also ﬂourished professional growth opportu-

nities after challenges had been overcome. This view

matches prior research where autonomy has been re-

ported to have a positive impact on job satisfaction

and as a general motivator for software developers,

while job satisfaction has been found to have a sig-

niﬁcant impact on turnover intentions as workers who

have high job satisfaction are less likely to leave (Lin

et al., 2017; Bass et al., 2018).

Nevertheless, lack of motivation has been empha-

sized as a correlated factor with reported intentions to

leave by prior research related to turnover among IS

professionals (Bass et al., 2018) as tapping into the in-

trinsic motivation needs of the software engineer cor-

relates to desirable outputs such as low staff turnover,

higher productivity, and better quality software (Noll

et al., 2017).

Our work also shows that autonomy received in an

early career has been stated to depend on support from

more experienced engineers or leadership, and this is

compliant with recent works where lack of supervi-

sor/management support is perceived as an inﬂuence

on software engineer’s turnover and dissatisfaction

(Uzoka et al., 2011; Robert Jr and You, 2018). Also,

autonomy is encouraged once this management sup-

port is facilitated effectively, fostering an environment

of trust and a culture of valuing individuals (Marinho

et al., 2021). Yet, organizational support with guid-

ance and certiﬁcation programs signiﬁcantly reduces

turnover intention (Chaves et al., 2022).

6 LIMITATIONS

Our study has some limitations that will be presented

in this section. We used a qualitative perspective

based on perceptions from participants, and it’s im-

portant to be aware of different contexts, environ-

ments, and individuals. Therefore, all data analysis

was based on this evidence, and we cannot state all

results match the full picture regarding autonomy and

turnover and distributed software development teams.

Furthermore, all data and classiﬁcations presented in

this paper must be treated carefully, as we only pro-

vide indications.

Our survey was applied to 181 participants over

two months, revealing time as a limitation. We en-

courage new research to take place during a longer

period. There was no deep statistical evaluation with

known models to address the relationship between the

cited factors in this paper. Therefore, we also suggest

this approach be taken in further studies.

7 CONCLUSIONS

Our study focuses on the relationship between auton-

omy and turnover within Distributed Software De-

velopment (DSD) teams. We have gathered valu-

able insights from real software engineers working

in DSD teams. Our study ﬁndings conﬁrm that au-

tonomy is linked to turnover and turnover intentions,

with various factors inﬂuencing this relationship. Ef-

fective communication, collaboration, trust, recogni-

tion, and task balance inﬂuence how autonomy im-

pacts turnover among software engineers of different

experience levels.

For early-career software engineers, providing au-

tonomy can lead to feelings of anxiety, insecurity,

and fear, which may result in lower commitment and

increased turnover intentions. This highlights the

importance of support from more experienced engi-

neers, leaders, and managers in providing guidance

and mentorship. On the other hand, experienced soft-

ware engineers are better equipped to handle auton-

omy without experiencing adverse outcomes. How-

ever, task overloading becomes a concern for them,

as increased accountability may lead to stress and an

unbalanced working routine, potentially leading to a

turnover.

Regardless of experience level, the perception of

autonomy as beneﬁcial for software engineers is con-

tingent upon effective communication, collaboration,

and mutual trust among team members in the work

environment.

Recognition plays a crucial role in sustaining the

beneﬁts of autonomy. Companies should establish

practices to acknowledge and reward software engi-

neers for achieving goals through their granted auton-

omy. Recognition, particularly when tied to salary

increases or bonuses, is valuable and can reduce

turnover intentions.

Lastly, our study highlights the perceived out-

comes for projects or companies when one or more

software engineers leave. Loss of knowledge and

decreased productivity were identiﬁed as the pri-

mary consequences. Further investigations are rec-

ommended to understand and address these issues.

In conclusion, our study sheds light on the re-

lationship between autonomy and turnover in DSD

teams, emphasizing the signiﬁcance of communica-

tion, collaboration, trust, recognition, and balancing

task responsibilities. Understanding and effectively

managing autonomy can create a positive work envi-

ronment and reduce turnover intentions among soft-

ware engineers.

KMIS 2023 - 15th International Conference on Knowledge Management and Information Systems

140

REFERENCES

Bao, L., Xing, Z., Xia, X., Lo, D., and Li, S. (2017). Who

will leave the company?: a large-scale industry study

of developer turnover by mining monthly work report.

In 2017 IEEE/ACM 14th International Conference on

Mining Software Repositories (MSR), pages 170–181.

IEEE.

Bass, J. M., Sarah, B., Razzak, M. A., and Noll, J. (2018).

Employee retention and turnover in global software

development: Comparing in-house offshoring and off-

shore outsourcing. In 2018 IEEE/ACM 13th Inter-

national Conference on Global Software Engineering

(ICGSE), pages 77–86. IEEE.

Chatzipetrou, P.,

Smite, D., and Van Solingen, R. (2018).

When and who leaves matters: emerging results from

an empirical study of employee turnover. In Pro-

ceedings of the 12th ACM/IEEE International Sym-

posium on Empirical Software Engineering and Mea-

surement, pages 1–4.

Chaves, A., Amorim, L., Marinho, M., de Farias J

unior, I.,

and Moura, H. (2022). Autonomy and turnover in dis-

tributed software development projects: a systematic

literature review. In Anais do XXV Congresso Ibero-

Americano em Engenharia de Software, pages 105–

119. SBC.

Dysvik, A. and Kuvaas, B. (2013). Perceived job autonomy

and turnover intention: The moderating role of per-

ceived supervisor support. European Journal of Work

and Organizational Psychology, 22(5):563–573.

Ebert, C., Kuhrmann, M., and Prikladnicki, R. (2016).

Global software engineering: Evolution and trends.

In 2016 IEEE 11th International Conference on

Global Software Engineering (ICGSE), pages 144–

153. IEEE.

Fitzgerald, B., Stol, K.-J., Min

or, S., and Cosmo, H. (2017).

Scaling a software business: the digitalization jour-

ney. Springer Nature.

Hackman, J. R. and Oldham, G. R. (1976). Motiva-

tion through the design of work: Test of a theory.

Organizational behavior and human performance,

16(2):250–279.

Hynninen, P., Piri, A., and Niinim

aki, T. (2010). Off-site

commitment and voluntary turnover in gsd projects.

In 2010 5th IEEE International Conference on Global

Software Engineering, pages 145–154. IEEE.

Lin, B., Robles, G., and Serebrenik, A. (2017). Developer

turnover in global, industrial open source projects:

Insights from applying survival analysis. In 2017

IEEE 12th International Conference on Global Soft-

ware Engineering (ICGSE), pages 66–75. IEEE.

Linaker, J., Sulaman, S. M., H

ost, M., and de Mello, R. M.

(2015). Guidelines for conducting surveys in software

engineering v. 1.1. Lund University, 50.

Lundene, K. and Mohagheghi, P. (2018). How autonomy

emerges as agile cross-functional teams mature. In

Proceedings of the 19th International Conference on

Agile Software Development: Companion, pages 1–5.

Marinho, M., Amorim, L., Camara, R., Oliveira, B. R., So-

bral, M., and Sampaio, S. (2021). Happier and further

by going together: The importance of software team

behaviour during the covid-19 pandemic. Technology

in society, 67:101799.

Marinho, M., Noll, J., and Beecham, S. (2018). Uncer-

tainty management for global software development

teams. In 11th International Conference on the Qual-

ity of Information and Communications Technology,

pages 238–246, Coimbra, Portugal. IEEE.

Marinho, M., Noll, J., Richardson, I., and Beecham, S.

(2019). Plan-driven approaches are alive and kick-

ing in agile global software development. In Interna-

tional Symposium on Empirical Software Engineering

and Measurement (ESEM), pages 1–11, Porto de Gal-

inhas, Brazil. IEEE.

Massoni, T., Ginani, N., Silva, W., Barros, Z., and Moura,

G. (2019). Relating voluntary turnover with job char-

acteristics, satisfaction and work exhaustion-an initial

study with brazilian developers. In 2019 IEEE/ACM

12th International Workshop on Cooperative and Hu-

man Aspects of Software Engineering (CHASE), pages

85–88. IEEE.

Moe, N. B.,

Smite, D., Paasivaara, M., and Lassenius, C.

(2021). Finding the sweet spot for organizational

control and team autonomy in large-scale agile soft-

ware development. Empirical Software Engineering,

26(5):101.

Noll, J., Beecham, S., Razzak, A., Richardson, B., Bar-

comb, A., and Richardson, I. (2017). Motivation and

autonomy in global software development. In Inter-

national Workshop on Global Sourcing of Informa-

tion Technology and Business Processes, pages 19–38.

Springer.

Remus, U., Wiener, M., Saunders, C., M

ahring, M., and

Koﬂer, M. (2016). Control modes versus control

styles: Investigating isd project control effects at the

individual level.

Robert Jr, L. P. and You, S. (2018). Are you satisﬁed yet?

shared leadership, individual trust, autonomy, and sat-

isfaction in virtual teams. Journal of the association

for information science and technology, 69(4):503–

513.

Uzoka, F.-M. E., Mgaya, K. V., Shemi, A. P., Kitindi, E. G.,

and Akinnuwesi, B. A. (2011). Stay or quit: It per-

sonnel turnover in botswana. In Proceedings of the

49th SIGMIS annual conference on Computer person-

nel research, pages 97–106.

Yener, S., Arslan, A., and Kilinc¸, S. (2020). The moderating

roles of technological self-efﬁcacy and time manage-

ment in the technostress and employee performance

relationship through burnout. Information Technology

& People.

APPENDIX - QUESTIONNAIRE

The applied survey questions and statements can be

accessed through this link: https://bit.ly/437WqYY.

Autonomy and Turnover: A Survey Applied to Distributed Software Teams

141