ArticlePDF Available

Abstract and Figures

Gamification is increasingly becoming a pertinent aspect of any UI and UX design. However, a canonical dearth in research and application of gamification has been related to the role of individual differences in susceptibility to gamification and its varied designs. To address this gap, this study reviews the extant corpus of research on tailored gamification (42 studies). The findings of the review indicate that most studies on the field are mostly focused on user modeling for a future personalization, adaptation, or recommendation of game elements. This user model usually contains the users’ preferences of play (i.e., player types), and is mostly applied in educational settings. The main contributions of this paper are a standardized terminology of the game elements used in tailored gamification, the discussion on the most suitable game elements for each users’ characteristic, and a research agenda including dynamic modeling, exploring multiple characteristics simultaneously, and understanding the effects of other aspects of the interaction on user experience.
Content may be subject to copyright.
Contents lists available at ScienceDirect
International Journal of Human-Computer Studies
journal homepage: www.elsevier.com/locate/ijhcs
Tailored gamification: A review of literature
Ana Carolina Tomé Klock
a,b,1,
, Isabela Gasparini
c,2
, Marcelo Soares Pimenta
b
, Juho Hamari
a
a
Gamification Group, Faculty of Information Technology and Communication Sciences, Tampere University, Finland
b
Institute of Informatics, Federal University of Rio Grande do Sul (UFRGS), Brazil
c
Department of Computer Science, Santa Catarina State University (UDESC), Brazil
ARTICLE INFO
Keywords:
Gamification
Tailoring
Personalization
Adaptation
Recommendation
Systematic review
ABSTRACT
Gamification is increasingly becoming a pertinent aspect of any UI and UX design. However, a canonical dearth
in research and application of gamification has been related to the role of individual differences in susceptibility
to gamification and its varied designs. To address this gap, this study reviews the extant corpus of research on
tailored gamification (42 studies). The findings of the review indicate that most studies on the field are mostly
focused on user modeling for a future personalization, adaptation, or recommendation of game elements. This
user model usually contains the users’ preferences of play (i.e., player types), and is mostly applied in educa-
tional settings. The main contributions of this paper are a standardized terminology of the game elements used in
tailored gamification, the discussion on the most suitable game elements for each users’ characteristic, and a
research agenda including dynamic modeling, exploring multiple characteristics simultaneously, and under-
standing the effects of other aspects of the interaction on user experience.
1. Introduction
Gamification refers to transforming activities, systems, services,
products, or organizational structures to afford gameful experiences
(Hamari, 2019). Beyond how the system has been designed, individual
differences, the context of use and aspects of the task can play an im-
portant role in the formation of the resultant experience (Hamari et al.,
2018; Hassenzahl and Tractinsky, 2006). Therefore, it would be bene-
ficial for designers, researchers and users to better understand how e.g.
contextual factors, individual traits, personality, interests, demographic
factors may moderate and impact the experience individuals have when
interacting with gamified systems (Attali and Arieli-Attali, 2015;
Bittner and Shipper, 2014; Conaway and Garay, 2014; Gil et al., 2015;
Pedro et al., 2015). Thus, tailoring the game elements according to the
users’ profile is a way to improve their experience while interacting
with a gamified system, and has been noted as a current trend in ga-
mification research (Koivisto and Hamari, 2019; Rapp et al., 2019).
In this sense, tailoring corresponds to any combination of in-
formation or change of strategy to reach individual needs and pre-
ferences according to one’s profile (Kreuter et al., 2013). Some concepts
(e.g., personalization, adaptation, and recommendation) can be re-
introduced in gamified scenarios to promote this tailoring effect. Per-
sonalization is a method where “the content is tailored by the system to
individual tastes” (Sundar and Marathe, 2010), while adaptation is a
way to “tailor interaction to different users in the same context”
(Brusilovsky and Maybury, 2002). Thus, both adaptive and persona-
lized systems modify some of their aspects to fulfill the specific user’s
needs with the most suitable solution (Garcia-Barrios et al., 2005). The
“user model” is the basis for those changes, since it stores explicitly and
implicitly captured data (e.g., goals, needs, preferences, and intentions)
(Jrad et al., 2007). However, while personalization is a type of adap-
tation that responds solely to the user model (Garcia-Barrios et al.,
2005), adaptation techniques also consider other models, such as the
domain, task, and discourse ones (Maybury, 1998). Recommendation
technique also reuses this user model, given that “recommendations are
usually personalized, different users [...] benefit from diverse, tailored
suggestions” (Ricci et al., 2015).
Based on these definitions, this paper aims to describe the state-of-
the-art of the tailored gamification through a systematic literature re-
view. We investigate methods, users’ characteristics, application
https://doi.org/10.1016/j.ijhcs.2020.102495
Received 2 September 2019; Received in revised form 8 June 2020; Accepted 12 June 2020
Corresponding author. fax: 5547999282858.
E-mail addresses: actklock@inf.ufrgs.br (A.C.T. Klock), isabela.gasparini@udesc.br (I. Gasparini), mpimenta@inf.ufrgs.br (M.S. Pimenta),
juho.hamari@tuni.fi (J. Hamari).
1
I am grateful for the financial support of CNPq - National Council for Scientific and Technological Development, and of the Coordenação de Aperfeiçoamento de
Pessoal de Nível Superior - Brasil (CAPES) – Finance Code 001.
2
I would like to thank the partial financial support of FAPESC, public call FAPESC/CNPq N
o
. 06/2016 support the infrastructure of CTI for young researchers,
project T.O. N
o
.: 2017TR1755 - Ambientes Inteligentes Educacionais com Integração de Técnicas de Learning Analytics e de Gamificação.
International Journal of Human-Computer Studies 144 (2020) 102495
Available online 13 June 2020
1071-5819/ © 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T
contexts, as well as the most suitable game elements for each user
profile according to extant corpus. For this, the paper is structured as
follows: Section 2 briefly presents the related systematic works about
tailored gamification, while introducing our major contributions.
Section 3 describes the systematic protocol and process, including the
research questions, keywords, search engines, selection criteria, and
extracted data. Section 4 discusses the quantitative and qualitative data
and answers the research questions, while Section 5 suggests a research
agenda. Lastly, Section 6 states the limitations and final remarks of this
study.
2. Related works
Several publications focusing on developing user preference models,
classifying game elements, or understanding the effects of different
tailoring approaches began to appear in recent years (Tondello, 2019).
It is widely accepted that there is a dearth of understanding of the role
of individual differences in the susceptibility of the effects of gamifi-
cation (Koivisto and Hamari, 2019), review studies approaching this
emergent topic to understand the literature better were expected.
For instance, Böckle et al. (2017) reviewed gamification literature
to answer “What are the main objectives, elements, and challenges of cur-
rent research regarding the development of adaptive gamification ap-
proaches?”. The 43 works included aimed to change the state of the user
(i.e., improving behavior and goals), support participation and
learning, or create meaning by adapting feedback and points based
mostly on usage data (e.g., performance or behavior) and user profile
(e.g., player types or personality traits). However, understanding the
effect of game elements on different individuals remains mostly un-
discussed, which makes it a current research challenge (Böckle et al.,
2017). Based on that, this study majorly contributes to the literature by
addressing the suitable game elements for the specific characteristics of
the user profile.
A more recent literature review conducted by Hallifax et al. (2019a)
aimed to answer “What are the current kinds of contributions to the field?”,
“What do the current contributions base their adaptation on, and what is the
effect of this adaptation on the gamified system?”, and “What is the impact
of the adaptive gamification, and how is this impact measured?”. The 20
studies included proposed real-world adaptation studies, re-
commendations, or theoretical architectures based on usage data and
user profile, similar to those reported by Böckle et al. (2017). Also,
Hallifax et al. (2019a) reported that studies lasting less than two weeks
were more likely to have positive effects than more extended experi-
ments, which described mitigated results. Since the scope of this study
was related to the educational area, the results are highly context-de-
pendent. Thus, our contribution focuses on exploring these tailoring
methods and outcomes regardless of the application context, given it is
a challenge for any domain (Cursino et al., 2018; Hakak et al., 2019;
Larson, 2020; Noorbehbahani et al., 2019).
3. Methodology
A systematic literature review is a method that analyzes the litera-
ture available of a specific phenomenon of interest, focusing on pro-
viding a background for new studies, identifying gaps for further in-
vestigation, and summarizing evidence concerning a technology
(Kitchenham and Charters, 2007). While tailoring gamification is an
emerging trend aiming to improve user experience by considering one’s
profile, researchers have not yet appropriately identified and sum-
marized the employed approaches and most suitable game elements to
do that. Following the guidelines previously described by
Kitchenham et al. (2009), this study aims to achieve it through a de-
fined, reproducible protocol describing: (1) research questions, (2)
keywords, (3) search engines, and (4) selection criteria.
3.1. Research questions
Four main research questions (RQ) will be answered to describe the
state-of-the-art of tailored gamification:
RQ1: What methods have been mainly used to make tailored game
elements in gamified systems based on the characteristics of the user
profile?
RQ1.1: Which algorithms or techniques are used to provide tai-
lored gamification?
RQ1.2: Which game elements are used in tailored gamification?
RQ1.3: How is tailored gamification evaluated?
RQ1.4: What aspects of gamification do these evaluations assess?
RQ2: In what contexts is the users’ profile considered for the tai-
lored gamification process?
RQ3: What users’ profile characteristics are considered?
RQ4: What are the most suitable game elements for each specific
characteristic?
3.2. Keywords
One of the most used processes to define the search keywords is
PICO, which identifies the population (P), the intervention (I), the
comparison (C), and the expected outcomes (O), based on the research
questions (Kitchenham and Charters, 2007). In this scenario, we have:
Population (P): studies that describe or apply gamification tailored
to characteristics of the user profile in computational systems;
Intervention (I): methods used to make tailored game elements in
gamified systems (e.g., adapted, personalized, recommended);
Comparison (C): not applicable, since the purpose of this study is to
describe the state-of-the-art;
Outcomes (O): most used algorithms, techniques, characteristics of
the user profile, and game elements.
A set of keywords representing the expected search results was
defined based on this PICO. Thus, the search string comprises two major
sets of keywords:
Gamification: covers the collection of keywords related to gamifi-
cation (e.g., gamified, gamify, gamifying);
Tailoring: covers the collection of keywords about tailor-made
methods (e.g., tailoring, tailored, adaptation, adaptive, user model,
user modeling, personalization, personalisation, recommend, re-
commender system, recommendation).
Thus, the search string defined is: “gamif* AND (adapt* OR model
OR personali* OR recommend* OR tailor*)” - where * is a wildcard.
3.3. Search engines
The search engines were defined based on the related studies
(Böckle et al., 2017; Hallifax et al., 2019a) and other systematic map-
ping and reviews about gamification, such as Dicheva et al. (2015);
Pedreira et al. (2015);de Sousa Borges et al. (2014) and Alahäivälä and
Oinas-Kukkonen (2016). All of these studies used ACM Digital Library
and IEEE Xplore, and four out of them used Science Direct, Scopus, and
Springer Link. Thus, these five search engines were also used in this
study.
3.4. Selection criteria
After searching for the keywords in the search engines, all 3400
studies returned in September 2018 were screened according to the
selection criteria (SC) to make the results more accurate.
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
2
SC1: Studies published in 2013 or later (last five years);
SC2: Studies written in English;
SC3: Full studies (with six or more pages);
SC4: Primary studies (i.e., not surveys, meta-analysis, systematic
mappings or reviews);
SC5: Studies available for download;
SC6: Non-duplicate studies (i.e., with the same Digital Object
Identifier - DOI);
SC7: Studies whose central theme is gamification;
SC8: Studies that explore any tailor-made methods;
SC9: Studies that propose or analyze the tailor-made game elements
in gamified systems.
Once applied in the returned studies, only 34 of them met the se-
lection criteria, as shown in Table 1. These studies (Akasaki et al., 2016;
AL-Smadi, 2015; Auvinen et al., 2015; Barata et al., 2016; Berg and
Petersen, 2013; Borges et al., 2016; Busch et al., 2016; Butler, 2014;
Challco et al., 2014; Denden et al., 2017a; 2017b; Fernandes and
Junior, 2016; Ferro et al., 2013; Fuß et al., 2014; Harteveld and
Sutherland, 2017; Herbert et al., 2014; Holmes et al., 2015; Lavoué
et al., 2018; Monterrat et al., 2015a; 2015b; Orji et al., 2017; 2014;
Oyibo et al., 2017a; 2017c; Paiva et al., 2015; Roosta et al., 2016;
Taspinar et al., 2016; Tondello et al., 2017a; 2017b; 2016; Utomo and
Santoso, 2015; Čudanov et al., 2014; Xu and Tang, 2015) have gone
through the backward snowballing process, which uses the reference
list of each one to identify additional work (Wohlin, 2014). From the
1720 references of the selected studies, eight other works met the se-
lection criteria and were also included in the systematic review analysis
(Codish and Ravid, 2014; 2017; Hakulinen and Auvinen, 2014; Jia
et al., 2016; Monterrat et al., 2014; 2017; Orji, 2014; Oyibo et al.,
2017b).
After generating association rules using the apriori algorithm
(Borgelt and Kruse, 2002) based on the titles, abstracts, and keywords
of the 42 selected works, it is possible to affirm that studies that con-
tained gamification-related words also included tailoring-related terms
(confidence coefficient = 1, support coefficient = 0.6). This result in-
dicates that both the keywords and the selection criteria were correctly
defined and applied.
3.5. Data to be extracted
Besides extracting the tailoring methods (RQ1), algorithms and
techniques (RQ1.1), game elements (RQ1.2), evaluation methodology
(RQ1.3), outcomes (RQ1.4), application contexts (RQ2), users’ char-
acteristics (RQ3) and the suitable game elements (RQ4), other quanti-
tative data were also collected: publication year, type of publication
venue (e.g., conference, journal) and author’s affiliation country.
4. Results and discussion
The interest of the researchers about tailored gamification has been
growing over the last years, with a substantial increase in the number of
studies between 2015 and 2017, as Fig. 1 illustrates. The publication
decreas in 2018 cannot be considered a trend since this search was
conducted before the end of the year. The subject is mostly published in
events (e.g., conferences, symposiums, workshops), which can be re-
lated to the longer revision time of journals and the low maturity of the
topic (due to its novelty).
According to the authors’ affiliation countries, Canadian and French
studies were the most published ones, as shown in Fig. 2, and many
international collaborations happened. Austrian, Canadian, and British
universities wrote collaboratively one of these studies (Tondello et al.,
2016) (each country had 0.3 paper – for calculation purposes). Another
four studies also shared the authorship between two countries, namely:
Austrian and Canadian (Busch et al., 2016), Canadian and French
(Monterrat et al., 2015a), Canadian and Spanish (Tondello et al.,
2017a), and Brazilian and Japanese (Borges et al., 2016) (0.5 paper
each). All the other 37 studies had no international collaboration.
4.1. Tailoring methods (RQ1)
Most of the studies (67%) focused on predicting and modeling the
user profile and the correspondent game elements, as shown in Table 2.
While user modeling is a very significant step towards tailored gami-
fication, this result indicates that researchers are still discussing what
instead of how to customize. Little has yet been explored about concrete
implementations of gamification tailoring methods (i.e., personaliza-
tion, adaptation, and recommendation strategies, according to what
each study described), implying a high possibility to expand and im-
prove this topic.
Table 1
Conduction of the Search.
Search engines
Criteria ACM Digital Library IEEE Xplore Scopus Science Direct SpringerLink Total
Search: September, 2018 172 271 1368 84 1505 3400
SC1: Published in 2013 or later 170 269 1359 84 1496 3378
SC2: Written in English 168 262 1317 83 1485 3315
SC3: Full studies 100 160 1124 79 1367 2830
SC4: Primary studies 96 154 923 76 1279 2528
SC5: Available for download 90 146 297 66 1276 1875
SC6: Non-duplicate studies 89 144 219 66 1188 1706
SC7: Gamification 73 113 175 38 372 771
SC8: Tailoring methods 26 29 50 3 84 192
SC9: Tailored gamification 10 8 8 1 7 34
Snowballing 8 42
Fig. 1. Publication years and Publication venue types.
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
3
4.1.1. Algorithms and techniques (RQ1.1)
The most used algorithms and techniques for tailored gamification
are exposed in Table 3. Given that studies were mainly related to user
modeling, the most used algorithms and techniques were evidently
from the statistics domain, aiming to correlate and test the most ap-
propriate game elements to each user profile dimension. The central
tendencies (e.g., mean, median), hypothesis tests (e.g., t-test, Wilcoxon
rank-sum test, Chi-squared test), deviation and dispersion (e.g., stan-
dard deviation, standard error), reliability (e.g., Cronbach’s alpha,
McDonalds’s omega, Cohen’s-kappa coefficient), linear regression
(Partial Least Squares, logistic regression), analysis of variance (e.g.,
One-way analysis of variance, Kruskal-Wallis test), and correlations
(e.g., Spearman’s rho, Pearson’s r) were the most employed techniques
within statistics area, while bayesian statistics (Barata et al., 2016) and
frequentist inference (Lavoué et al., 2018) also appeared once. How-
ever, solutions are also emerging from other domains, such as in-
formation sciences (e.g., manual classification, decision tree, frame-
work (Čudanov et al., 2014), and ontology (Challco et al., 2014)),
research (e.g., literature review and qualitative analysis), adaptation
(e.g., adaptation rules, and user profiling (Monterrat et al., 2017)),
recommendation (e.g., recommender systems (Tondello et al., 2017b;
Xu and Tang, 2015)), data mining (e.g., clustering (Hakulinen and
Auvinen, 2014; Herbert et al., 2014), causal data mining (Xu and
Tang, 2015), and Receiver Operating Characteristic (Barata et al.,
2016)), artificial intelligence (e.g., pedagogical agent (Utomo and
Santoso, 2015)), machine learning (e.g., correlation-based feature
selection, expectation-maximization, k-nearest neighbors and sequen-
tial minimal optimization (Barata et al., 2016)), human-computer
interaction (e.g., design of motivational affordances (Harteveld and
Sutherland, 2017)), and linear algebra (e.g., matrix multiplication
(Monterrat et al., 2017)).
4.1.2. Game elements (RQ1.2)
Gamification employs several game elements to get a meaningful
response from users (Zichermann and Cunningham, 2011), regardless
of the application context. However, the literature uses different ter-
minologies to define a common game element, since some works em-
ploy definitions of distinct abstraction levels (e.g., progression and level
could address the same element, depending on the work). While this is a
well-known issue for the overall gamification subject (Hallifax et al.,
2019b), the following game elements emerged from the included stu-
dies, which was standardized based on the authors’ description or
image of each game element defined:
Anarchy: creates an environment without any restrictions or pe-
nalties (Butler, 2014), allowing anarchic gameplay (Tondello et al.,
2016);
Anonymity: is the opportunity to share the data (e.g., performance,
opinions) in the system without naming the user (Tondello et al.,
2016);
Badge: is a visual representation of the user’s achievements. It is
usually described as “badges” (Akasaki et al., 2016; Barata et al.,
2016; Butler, 2014; Challco et al., 2014; Codish and Ravid, 2014;
Denden et al., 2017a; 2017b; Ferro et al., 2013; Hakulinen and
Auvinen, 2014; Herbert et al., 2014; Jia et al., 2016; Paiva et al.,
2015; Roosta et al., 2016; Tondello et al., 2017a; 2016; Utomo and
Santoso, 2015; Čudanov et al., 2014) or “achievements” (Auvinen
et al., 2015; Berg and Petersen, 2013; Borges et al., 2016; Fernandes
and Junior, 2016; Fuß et al., 2014);
Challenge: can be a variety of situations to deal with or figure them
out (Butler, 2014), boss battles (Holmes et al., 2015), or any other
kind of action that requires effort from the user to be completed.
Authors usually describe this game element as challenge (AL-Smadi,
Fig. 2. Countries of the authors’ affiliations.
Table 2
Tailored gamification methods.
Tailoring method Studies Total
User modeling Akasaki et al. (2016);AL-Smadi (2015);Auvinen et al. (2015);Barata et al. (2016);Berg and Petersen (2013);Borges et al. (2016);
Busch et al. (2016);Butler (2014);Codish and Ravid (2014, 2017);Denden et al. (2017a,b);Fernandes and Junior (2016);Fuß et al. (2014);
Hakulinen and Auvinen (2014);Harteveld and Sutherland (2017);Herbert et al. (2014);Holmes et al. (2015);Jia et al. (2016);Orji (2014);
Orji et al. (2017);Oyibo et al. (2017a, 2017b, 2017c);Taspinar et al. (2016);Tondello et al. (2017a, 2016);Čudanov et al. (2014)
28 (67%)
Personalization Challco et al. (2014);Ferro et al. (2013);Orji et al. (2018, 2014);Paiva et al. (2015);Roosta et al. (2016);Utomo and Santoso (2015) 7(16%)
Adaptation Lavoué et al. (2018);Monterrat et al. (2015a, 2014, 2015b, 2017) 5(12%)
Recommendation Tondello et al. (2017b);Xu and Tang (2015) 2(4%)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
4
Table 3
Algorithms and techniques used by each tailoring method.
Statistics Inf. S. Research Ad.
Studies Central
tendency
Hypothesis
test
Deviation
and
Dispersion
Reliability Linear
regression
Analysis
of
variance
Correlation Factor
Analysis
Structural
Equation
Models
Principal
Component
Analysis
Manual
classification
Decision
tree
Literature
review
Qualitative
analysis
Adaptation
rules
User modeling Berg and Petersen (2013)
Butler (2014) • •
Codish and Ravid (2014) • •
Fuß et al. (2014) • •
Hakulinen and
Auvinen (2014)
• •
Herbert et al. (2014)
Orji (2014) • •
AL-Smadi (2015)
Auvinen et al. (2015) • •
Holmes et al. (2015)
Akasaki et al. (2016)
Barata et al. (2016) • •
Borges et al. (2016)
Busch et al. (2016) • •
Fernandes and
Junior (2016)
Jia et al. (2016) • •
Taspinar et al. (2016)
Tondello et al. (2016) • •
Codish and Ravid (2017) • •
Denden et al. (2017a) • •
Denden et al. (2017b) • •
Orji et al. (2017) • •
Oyibo et al. (2017a) • •
Oyibo et al. (2017b) • •
Oyibo et al. (2017c)
Tondello et al. (2017a) • •
Personaliz. Ferro et al. (2013) • •
Orji et al. (2014) • • • •
Roosta et al. (2016)
Orji et al. (2018) • • • •
Adaptation Monterrat et al. (2014)
Monterrat et al. (2015a) • •
Monterrat et al. (2015b)
Monterrat et al. (2017)
Lavoué et al. (2018) • •
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
5
2015; Barata et al., 2016; Berg and Petersen, 2013; Butler, 2014;
Harteveld and Sutherland, 2017; Jia et al., 2016; Monterrat et al.,
2014; 2015b; Taspinar et al., 2016; Tondello et al., 2017a; 2016),
quest (Borges et al., 2016; Challco et al., 2014; Fernandes and
Junior, 2016; Ferro et al., 2013; Herbert et al., 2014), or task (Fuß
et al., 2014);
Choice: lets the user have the autonomy to determinate his/her
verdict among many possibilities (Harteveld and Sutherland, 2017).
This game element includes non-linear gameplay (Tondello et al.,
2016), high-level closures (Holmes et al., 2015), numerous open-
ended or branching decisions to make and multiple paths to choose
(AL-Smadi, 2015; Butler, 2014), and anchor juxtaposition (i.e., the
user must decide between paying for more time or making more
efforts to achieve the same result) (Tondello et al., 2017a);
Collection: creates the sense of ownership in the system through
picking-up (Holmes et al., 2015), trading (Tondello et al., 2017a;
2016), and management collectibles and resources within an in-
ventory (Akasaki et al., 2016; AL-Smadi, 2015; Butler, 2014; Ferro
et al., 2013; Fuß et al., 2014; Monterrat et al., 2014; 2015b);
Competition: allows social comparison (AL-Smadi, 2015; Berg and
Petersen, 2013; Busch et al., 2016; Holmes et al., 2015; Orji, 2014;
Orji et al., 2017; 2018; 2014; Oyibo et al., 2017a; 2017b; 2017c;
Tondello et al., 2017a; 2016) between users and promotes a possi-
bility to prove themselves better than others. This competition en-
compasses personal conflicts (Berg and Petersen, 2013; Monterrat
et al., 2014; 2015b), duels (Fuß et al., 2014), and contests
(Butler, 2014);
Consequence: facilitates the user observation of the cause-and-ef-
fect linkage of his/her behavior (Busch et al., 2016; Orji, 2014; Orji
et al., 2017; 2018; 2014). This element also encompasses the set of
boundaries (Monterrat et al., 2014; 2015b) or rules (Butler, 2014)
and the punishment for not following it (Orji et al., 2017; 2018);
Customization: enables the self-expression of the users through the
creation and decoration of their virtual space (AL-Smadi, 2015; Berg
and Petersen, 2013; Monterrat et al., 2014; 2015b), their avatar
(Borges et al., 2016; Butler, 2014; Denden et al., 2017a; 2017b; Jia
et al., 2016; Tondello et al., 2017a) or their character (Harteveld
and Sutherland, 2017; Holmes et al., 2015; Monterrat et al., 2014;
2015b), and the personalization (Busch et al., 2016; Challco et al.,
2014; Fernandes and Junior, 2016; Ferro et al., 2013; Fuß et al.,
2014; Herbert et al., 2014; Orji, 2014; Orji et al., 2017; 2018; 2014;
Tondello et al., 2016) of some aspects of the system’s interface;
Easter egg: is a surprising response of the system to a specific action
of the user (Tondello et al., 2017a; 2016). This element can appear
as a gift of surprise (Čudanov et al., 2014), an unforeseen event
(Butler, 2014), or a secret lesson (Fuß et al., 2014);
Emotion: involves all aesthetics (Monterrat et al., 2014; 2015b) that
create positive perceptions (e.g., enjoyment (Butler, 2014), play-
fulness (Codish and Ravid, 2014)) on users and influence their ac-
tions (Harteveld and Sutherland, 2017);
Exploration: gives the user the possibility to investigate and dis-
cover areas and features of the system, through exploratory tasks
(AL-Smadi, 2015; Butler, 2014; Tondello et al., 2016), imperfect
information (Holmes et al., 2015), and mystery boxes
(Tondello et al., 2017a);
Feedback: returns relevant information to users. While some au-
thors directly described it as “feedback” (Berg and Petersen, 2013;
Denden et al., 2017a; 2017b; Harteveld and Sutherland, 2017; Jia
et al., 2016; Orji et al., 2018; Roosta et al., 2016; Taspinar et al.,
2016), others employed feedback to provide a way for users self-
monitor themselves (Busch et al., 2016; Orji, 2014; Orji et al., 2017;
2018; 2014) or a game state overview (AL-Smadi, 2015; Holmes
et al., 2015);
Gifting: allows the user to give or share resources with others (AL-
Smadi, 2015), encouraging altruism and satisfaction (Challco et al.,
2014; Herbert et al., 2014; Tondello et al., 2017a);
Guild: includes the team feeling (i.e., being part of a group) and the
caretaking between users. Other than being directly called “guild”
or “team” (Tondello et al., 2017a; 2016), it can appear as co-
operation (Busch et al., 2016; Fuß et al., 2014; Orji, 2014; Orji et al.,
2017; 2018; 2014), collaboration (AL-Smadi, 2015; Berg and
Petersen, 2013; Taspinar et al., 2016), and support (Butler, 2014;
Fuß et al., 2014) of the guildmates. In most of the applications, the
system creates roles for guild administrators to look after other users
(Tondello et al., 2017a);
Honor system: creates a reputation score where other users or even
the system calculates a number that indicates the user reliability
(Borges et al., 2016);
Leaderboard: orders the users according to some criteria (e.g.,
points, levels, badges), contextualizing the other game elements to
enable user comparison (Borges et al., 2016; Challco et al., 2014;
Codish and Ravid, 2014; 2017; Denden et al., 2017a; 2017b; Ferro
et al., 2013; Herbert et al., 2014; Jia et al., 2016; Lavoué et al.,
2018; Monterrat et al., 2015a; 2017; Roosta et al., 2016; Tondello
et al., 2017a; 2016). This element can also appear in the literature as
a ranking (Akasaki et al., 2016; Fernandes and Junior, 2016;
Taspinar et al., 2016);
Learning: allows the user to gain and master new skills (Butler,
2014; Fuß et al., 2014; Holmes et al., 2015; Monterrat et al., 2017),
by imitating other users (i.e., social learning) (Berg and Petersen,
2013; Oyibo et al., 2017a; 2017b; 2017c; Tondello et al., 2017a;
2016) or through visual representation (e.g., knowledge maps
(Borges et al., 2016), skill trees (Barata et al., 2016));
Level: supports users to track their progression through the system’s
purpose over time, aiding the visualization (e.g., progress bar
(Borges et al., 2016; Codish and Ravid, 2014; 2017; Denden et al.,
2017a; 2017b; Fernandes and Junior, 2016; Ferro et al., 2013;
Herbert et al., 2014; Holmes et al., 2015; Jia et al., 2016; Roosta
et al., 2016; Taspinar et al., 2016; Tondello et al., 2017a; 2016),
stars (Lavoué et al., 2018; Monterrat et al., 2017) or flags along a
path (Monterrat et al., 2015a)) of a continuously and gradually
growing towards a specific goal (Butler, 2014; Fuß et al., 2014);
Lottery: is an element of randomness (i.e., a chance) within the
system, such as offering a “free lunch” (Tondello et al., 2017a);
Meaning: allows the user to auto-identify with the system through a
common purpose (Holmes et al., 2015). Some examples of mean-
ingful choice are explicit objectives to pursue (Berg and Petersen,
2013; Butler, 2014; Jia et al., 2016; Monterrat et al., 2014; 2015b;
Taspinar et al., 2016), goal setting and suggestion (Orji et al., 2017;
2018), and actions tied to something more significant than the user
himself or herself (e.g., humanity hero) (Tondello et al., 2017a);
Narrative: includes plots that connect the other game elements.
Besides a linear or unfolding sequence of events (i.e., story) (Akasaki
et al., 2016; Butler, 2014; Monterrat et al., 2014; 2015b; Tondello
et al., 2017a), they can encompass different themes (Tondello et al.,
2017a) and contexts that make sense to the users (Butler, 2014),
plot-twists (Butler, 2014), dramatic art (Monterrat et al., 2014;
2015b), and life-like agents (Utomo and Santoso, 2015);
Point: is numerical feedback provided when the user does a specific
action (Borges et al., 2016; Challco et al., 2014; Codish and Ravid,
2014; 2017; Denden et al., 2017a; 2017b; Fernandes and Junior,
2016; Ferro et al., 2013; Herbert et al., 2014; Jia et al., 2016;
Taspinar et al., 2016; Tondello et al., 2017a; 2016). It is also called
“experience point” since it indicates that the user’s progress (and
knowledge) is continuously growing (Barata et al., 2016; Fuß et al.,
2014);
Prize: is any reward that the user wins for his/her action (Busch
et al., 2016; Codish and Ravid, 2014; Jia et al., 2016; Orji, 2014;
Orji et al., 2017; 2018; 2014; Oyibo et al., 2017a; 2017b; 2017c;
Tondello et al., 2016). Some examples of this game element are
bonuses (Fernandes and Junior, 2016; Ferro et al., 2013), combos
(Fernandes and Junior, 2016; Ferro et al., 2013), win states (AL-
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
6
Smadi, 2015; Fuß et al., 2014), and boosters (Tondello et al.,
2017a);
Reward schedule: applies reinforcements consistently to condition
and strengthens the user’s behavior in anticipation of new rewards.
While other authors defined it as “reward schedule” (Fernandes and
Junior, 2016; Ferro et al., 2013), Butler (2014) introduced it as
“infrequent but long sessions should be rewarding”;
Signposting: is any guidance that the system provides to the users
to help, suggest, or warn them about a path to be (or not) followed
(AL-Smadi, 2015; Busch et al., 2016). It appeared as a suggestion
(Busch et al., 2016; Utomo and Santoso, 2015), a tip (Lavoué et al.,
2018; Monterrat et al., 2015a), a greeting message (AL-
Smadi, 2015), a glowing choice (i.e., highlighting an item)
(Tondello et al., 2017a), or even as an intuitive play (Butler, 2014);
Single-player: allows the user to play alone, no requirements of
social interaction with others (Butler, 2014);
Social discovery: supports the user to find or be found by other
people with the same interests or status by checking the public
profile (Fuß et al., 2014; Tondello et al., 2017a; 2016);
Social network: enables the connection between users (Challco
et al., 2014; Herbert et al., 2014; Tondello et al., 2017a; 2016)
through communication channels (Holmes et al., 2015) that support
the human-human interaction (Harteveld and Sutherland, 2017)
(e.g., likes (AL-Smadi, 2015), chat (Denden et al., 2017a; 2017b),
voice, and face-to-face (Butler, 2014)). This communication can also
appear as multi-player games (Butler, 2014; Holmes et al., 2015);
Social pressure: permits users to influence or be influenced by
others (Harteveld and Sutherland, 2017), through public comments
(AL-Smadi, 2015) or creating fear of becoming “the loser” in the
team (e.g., allowing others to know how many time you have spent
on a specific level) (Tondello et al., 2017a);
Social status: allows the user to brag about himself/herself through
the system. Other than “social status” (Challco et al., 2014; Ferro
et al., 2013; Fuß et al., 2014; Herbert et al., 2014; Tondello et al.,
2017a), some types of implementation are praises (Busch et al.,
2016; Orji, 2014; Orji et al., 2014), certificates (Tondello et al.,
2017a; 2016), public ranks and titles (Butler, 2014), and the pos-
sibility to show off (AL-Smadi, 2015);
Strategy: requires the user to elaborate a plan of action to maximize
opportunities or minimizing losses and damages (Butler, 2014). It
can also be activities in which users review the action plan (AL-
Smadi, 2015) and notice a perceived chance to succeed
(Holmes et al., 2015);
Time pressure: involves a specific constraint that demands the user
to complete a particular task in a determined time. This game ele-
ment encompasses timers (Butler, 2014; Lavoué et al., 2018;
Monterrat et al., 2015a), and deadlines (Butler, 2014; Monterrat
et al., 2015a), countdowns (Taspinar et al., 2016; Utomo and
Santoso, 2015) and task reminders (Utomo and Santoso, 2015);
Unlockable: is an exclusive content conditioned to an action of the
user to be available, such as new contents (Herbert et al., 2014;
Lavoué et al., 2018; Tondello et al., 2017a; 2016) and features
(Challco et al., 2014; Holmes et al., 2015);
Virtual economy: creates a currency (Borges et al., 2016) to allow
users to purchase (AL-Smadi, 2015) features and virtual goods
(Borges et al., 2016; Challco et al., 2014; Herbert et al., 2014), and
using the term “virtual economy” is also common (Tondello et al.,
2017a; 2016);
Voting: allows the user to give his/her opinion within a subject,
including any rating and voting mechanism (AL-Smadi, 2015;
Tondello et al., 2017a; 2016).
Based on this terminology, the majority of the studies applied cus-
tomization (55%) and badges (52%) in user modeling, personalization,
adaptation, or recommendation processes. Other relevant game ele-
ments also used were: challenges and levels (43%), and competition
and leaderboards (40%). The two recommendation studies (Tondello
et al., 2017b; Xu and Tang, 2015) were the only ones that did not
specify the game elements considered. All game elements used in more
than 14% of the studies are represented in Fig. 3. Contrary to the
forecast from the early years of gamification, the PBL triad (Points,
Badges, and Leaderboards) is no longer the dominant set of applied
game elements, at least in the tailoring context. Instead, researchers
have been searching for the effects of a wide range of game elements in
the user experience.
4.1.3. Evaluation methodology (RQ1.3) and outcomes (RQ1.4)
A total of 9 studies did not perform any empirical evaluation,
mainly focusing on proposing game elements based on literature review
(Denden et al., 2017b; Ferro et al., 2013; Monterrat et al., 2015b) or
design principles (AL-Smadi, 2015; Fernandes and Junior, 2016;
Harteveld and Sutherland, 2017; Orji et al., 2017; Tondello et al.,
2017b; Čudanov et al., 2014). The other 33 studies applied a total of 54
evaluative methods, with a peak of using four methods simultaneously
Fig. 3. Most used game elements in the studies.
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
7
Table 4
Evaluation methods and outcomes.
Methods
Studies Survey Questionnaire System usage Storyboard Experiment Interview Personas Prototyping User observation
(Berg and Petersen, 2013) •
(Butler, 2014) •
(Challco et al., 2014) •
(Codish and Ravid, 2014) •
(Fuß et al., 2014) •
(Hakulinen and Auvinen, 2014) •
(Herbert et al., 2014) •
(Monterrat et al., 2014) •
(Orji, 2014) • •
(Orji et al., 2014) •
(Auvinen et al., 2015) •
(Holmes et al., 2015) •
(Monterrat et al., 2015a) •
(Paiva et al., 2015) •
(Utomo and Santoso, 2015) •
(Xu and Tang, 2015) •
(Akasaki et al., 2016) •
(Barata et al., 2016) • •
(Borges et al., 2016) •
(Busch et al., 2016) •
(Jia et al., 2016) •
(Roosta et al., 2016) • •
(Taspinar et al., 2016) •
(Tondello et al., 2016) •
(Codish and Ravid, 2017) •
(Denden et al., 2017a) •
(Monterrat et al., 2017) •
(Oyibo et al., 2017a) •
(Oyibo et al., 2017b) •
(Oyibo et al., 2017c) •
(Tondello et al., 2017a) •
(Lavoué et al., 2018) •
(Orji et al., 2018) •
Total 15 15 7 5 4 2 2 2 2
Outcomes
Studies Motivation Persuasion Interaction Enjoyment Performance Perception Playfulness Attitude Effectiveness
(Berg and Petersen, 2013) •
(Butler, 2014) • •
(Challco et al., 2014) •
(Codish and Ravid, 2014) • •
(Fuß et al., 2014) •
(Hakulinen and Auvinen, 2014) •
(Herbert et al., 2014) •
(Monterrat et al., 2014) •
(Orji, 2014) •
(Orji et al., 2014) •
(Auvinen et al., 2015) • •
(Holmes et al., 2015) •
(continued on next page)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
8
(Berg and Petersen, 2013). The most common methods for evaluating
tailored gamification were surveys and questionnaires (45% each), as
shown in Table 4. While a survey is a systematic method for gathering
information, which involves the design, data collection through ques-
tionnaires, processing, and analysis (Groves et al., 2011), a ques-
tionnaire is an instrument based on a set of questions, devised for a
survey or statistical study (Stevenson, 2010). The most prominent
samples were obtained through surveys (
=n6134,
giving almost 323
participants for each survey), storyboards (
=n3351,
an average of 838
participants for each study), and questionnaires (
=n1760,
a proportion
of 110 participants per study). Since all storyboard evaluations also
employed surveys, more than half of the participants of using this
technique derived from surveys. It further suggests that surveys and
questionnaires were the most used in the literature (both in studies
quantity and sample sizes), which is likely related to the facility to
achieve more significant samples with these methods.
The selected studies evaluated fourteen different gamification out-
comes, mostly concentrating on motivation (36%) and persuasion
(27%). The main difference between these two concepts is that, while
motivation is “the general desire or willingness to do something”,
persuasion is “a process designed to change the attitude or behaviour of
a person or group from their current view to a view that the persuader
wants them to hold” (Franklin, 2011). In other words, most studies are
evaluating how effective gamification is to encourage (i.e., motivate)
and convince (i.e., persuade) people. The sums of the sample sizes re-
ported by the original studies are also the greatest ones for persuasion
(
=n4569
) and motivation (
), demonstrating the pre-
dominance of these two evaluative aspects in both quantity and size.
Also, six studies covered more than one outcome (Akasaki et al., 2016;
Auvinen et al., 2015; Butler, 2014; Codish and Ravid, 2014; Holmes
et al., 2015; Orji, 2014).
4.2. Application contexts (RQ2)
Of the 42 papers, almost 60% of them applied the tailored gamifi-
cation in the educational context, as shown in Table 5. Another ten
studies did not focus on a specific context, claiming to be easily applied
to any. Other contexts, such as health, academia (i.e., publishing pa-
pers), ecology, government, and services, also appeared. Researchers
from the educational area mainly use Moodle (20% -
Barata et al. (2016);Codish and Ravid (2017);Denden et al. (2017a,b);
Utomo and Santoso (2015)) and Projet Voltaire (12% - Lavoué
et al. (2018);Monterrat et al. (2015a, 2017)) as the virtual learning
environment to implement the tailored gamification. Meanwhile, some
authors are proposing conceptual frameworks (12% -
Borges et al. (2016);Monterrat et al. (2014, 2015b)) and research
models (8% - Codish and Ravid (2014);Ferro et al. (2013)). In the
educational context, the most used game elements were badges, lea-
derboards, levels, challenges, customization, and points, which means
that the PBL triad is still popular in this area, but other game elements
also appeared.
Generic context is the only one in which the results are not from any
concrete, implemented gamification. Instead, all of the results came
from questionnaires (Busch et al. (2016);Jia et al. (2016);
Orji et al. (2017);Oyibo et al. (2017a,b)), research models (Orji (2014);
Oyibo et al. (2017c);Tondello et al. (2017a)), and conceptual frame-
works (Butler (2014);Tondello et al. (2017b)). Also, the most explored
game elements in the generic context were distinct from the educa-
tional one: competition, prizes, customization, guilds, and learning.
This comparison could lead us to two different conclusions: either i)
researchers are randomly naming correlated game elements (e.g.,
competition and leaderboards) due to the lack of proper terminology, or
ii) each context is going to the opposite path of the other, by using more
general or specific abstraction levels for the game elements.
In the health context, two out of three studies obtained their results
through questionnaires (Orji et al., 2018; 2014), while one study
Table 4 (continued)
Outcomes
Studies Motivation Persuasion Interaction Enjoyment Performance Perception Playfulness Attitude Effectiveness
(Monterrat et al., 2015a) •
(Paiva et al., 2015) •
(Utomo and Santoso, 2015) •
(Xu and Tang, 2015) •
(Akasaki et al., 2016) •
(Barata et al., 2016) •
(Borges et al., 2016) •
(Busch et al., 2016) •
(Jia et al., 2016) •
(Roosta et al., 2016) •
(Taspinar et al., 2016) •
(Tondello et al., 2016) •
(Codish and Ravid, 2017) •
(Denden et al., 2017a) •
(Monterrat et al., 2017) •
(Oyibo et al., 2017a) •
(Oyibo et al., 2017b) •
(Oyibo et al., 2017c) •
(Tondello et al., 2017a) •
(Lavoué et al., 2018) •
(Orji et al., 2018) •
Total 12 9 5 4 3 3 2 1 1
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
9
applied the tailored game elements in many interactive rehabilitation
systems (Holmes et al., 2015). All three studies explored competition,
customization, and feedback, while two of them applied consequences,
guilds, meaning, and prizes. Although the health context is still little
explored, we can observe that some game elements employed in the
generic context also appear in this one.
4.3. Users’ Characteristics (RQ3) and suitable game elements (RQ4)
The player type, which classifies users according to their game
preferences and play styles, was studied by 45% of the works, followed
by gender (14%) and personality traits (12%). While most of the studies
covered only one characteristic at the time, others also simultaneously
analyzed gamification effects in two (behavior and performance by
Utomo and Santoso (2015), behavior and gender by
Denden et al. (2017b), age and gender by Oyibo et al. (2017b)), three
(age, culture and gender in Oyibo et al. (2017a)), and even four (age,
gender, player typology and personality traits in
Tondello et al. (2017a)) different characteristics. Thus, a reasonable
part of the studies examining gender or culture were also covering other
characteristics, and no study analyzed performance exclusively. Also,
studies investigating player typology, personality traits, and motivation
used more than one theory, represented in Fig. 4. While the results of
these studies rely on the adopted methodological approaches, the game
elements suggested by the authors for each characteristic are detailed
below.
4.3.1. Player typologies
Among the studies exploring the effects of the game elements for
each player type, most of them employed the typology proposed by
Bartle. This typology defines four different player types based on a two-
dimensional interplay of play styles: world versus player, and action
versus iteration. These player types are achievers (acting upon the
world), explorers (interacting with the world), killers (acting upon other
players), and socializers (interacting with other players). In other words,
achievers enjoy reaching personal goals, explorers prefer knowing all
tricks and locations available in the game, killers need to impose
themselves on the others to feel fulfilled, and socializers like to talk and
interact with other players within the game. Each user is a composition
of player types, and one or some of them are usually predominant. The
system’s usage and questionnaires are techniques that can be applied to
identify the player types.
Seven studies correlated 28 different game elements with Bartle
typology. Three studies suggested challenges and levels for achievers;
collections for explorers; badges, leaderboards, and levels for killers;
and guilds for socializers, as shown in Table 6. Other game elements
appeared twice for achievers (badges, customization, and leader-
boards), explorers (badges, challenges, customization, exploration,
feedback, guilds, and points), killers (challenges, points, prizes, social
pressure and status), and socializers (customization, and social net-
work). At least one study from the literature supports that badges,
challenges, customization, emotions, feedback, levels, and meaning are
game elements appreciated by all of these player types. However, there
is a divergence in the number of game elements to each player type:
literature suggested 18 different game elements to achievers, 20 to
explorers, 16 to killers, and 13 to socializers. Several relevant points
remain unclear since the studies do not usually describe the distribution
of user’s characteristics from their sample, such as: Are socializers less
likely to appreciate the game elements, or were the game elements
explored by the studies just not employing the more favorable ones to
this player type? Given each user is a composition of different player
types, what is the impact of the less predominant player types in each
user’s preferences for the game elements mentioned above? Based on
this composition, should we exclusively look at the more predominant
one, or is there a “cutoff score” for the player type percentages that we
should consider?
Table 5
Contexts of tailored gamification applications.
Context Studies Total
Education AL-Smadi (2015);Auvinen et al. (2015);Barata et al. (2016);Berg and Petersen (2013);Borges et al. (2016);Challco et al. (2014);Codish and Ravid
(2014, 2017);Denden et al. (2017a,b);Ferro et al. (2013);Fuß et al. (2014);Hakulinen and Auvinen (2014);Harteveld and Sutherland (2017);
Herbert et al. (2014);Lavoué et al. (2018);Monterrat et al. (2015a, 2014, 2015b, 2017);Paiva et al. (2015);Roosta et al. (2016);
Taspinar et al. (2016);Tondello et al. (2016);Utomo and Santoso (2015)
25 (59.5%)
Generic Busch et al. (2016);Butler (2014);Jia et al. (2016);Orji (2014);Orji et al. (2017);Oyibo et al. (2017a, 2017b, 2017c);Tondello et al. (2017a,b) 10 (23.8%)
Health Holmes et al. (2015);Orji et al. (2018, 2014) 3(7.1%)
Academia Čudanov et al. (2014) 1(2.4%)
Ecology Xu and Tang (2015) 1(2.4%)
Government Fernandes and Junior (2016) 1(2.4%)
Services Akasaki et al. (2016) 1(2.4%)
Fig. 4. Users’ characteristics considered to tailor the game elements.
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
10
Table 6
Suggested game elements for each player type of Bartle typology.
Game element Achiever Explorer Killer Socializer
Badge 2 (Akasaki et al., 2016; Fernandes and Junior, 2016) 2 (Akasaki et al., 2016; Fuß et al., 2014) 3 (Akasaki et al., 2016; Fernandes and Junior, 2016;
Čudanov et al., 2014)
1 (Čudanov et al., 2014)
Challenge 3 (Fuß et al., 2014; Harteveld and Sutherland, 2017;
Taspinar et al., 2016)
2 (Fernandes and Junior, 2016; Taspinar et al.,
2016)
2 (Al-Smadi, 2015; Taspinar et al., 2016) 1 (Fernandes and Junior, 2016)
Choice 1 (AL-Smadi, 2015) 1 (Harteveld and Sutherland, 2017)
Collection 1 (Akasaki et al., 2016) 3 (Akasaki et al., 2016; AL-Smadi, 2015; Fuß
et al., 2014)
1 (Akasaki et al., 2016)
Competition 2 (AL-Smadi, 2015; Fuß et al., 2014)
Customization 2 (AL-Smadi, 2015; Fuß et al., 2014) 2 (Fuß et al., 2014; Harteveld and Sutherland,
2017)
1 (Fuß et al., 2014) 2 (Fernandes and Junior, 2016; Fuß et al., 2014)
Easter egg 1 (Čudanov et al., 2014) 1 (Fuß et al., 2014)
Emotion 1 (Harteveld and Sutherland, 2017) 1 (Harteveld and Sutherland, 2017) 1 (Harteveld and Sutherland, 2017) 1 (Harteveld and Sutherland, 2017)
Exploration 1 (Akasaki et al., 2016) 2 (Akasaki et al., 2016; AL-Smadi, 2015) 1 (Akasaki et al., 2016)
Feedback 2 (Harteveld and Sutherland, 2017; Taspinar et al., 2016) 2 (AL-Smadi, 2015; Taspinar et al., 2016) 1 (Taspinar et al., 2016) 1 (Taspinar et al., 2016)
Gifting 1 (AL-Smadi, 2015)
Guild 1 (Taspinar et al., 2016) 2 (AL-Smadi, 2015; Taspinar et al., 2016) 3 (Fuß et al., 2014; AL-Smadi, 2015; Taspinar
et al., 2016)
Leaderboard 2 (Akasaki et al., 2016; Taspinar et al., 2016) 1 (Akasaki et al., 2016) 3 (Akasaki et al., 2016; Fernandes and Junior, 2016;
Taspinar et al., 2016)
Learning 1 (Fuß et al., 2014)
Level 3 (Fernandes and Junior, 2016; Fuß et al., 2014; Taspinar
et al., 2016)
1 (Fuß et al., 2014) 3 (Fernandes and Junior, 2016; Fuß et al., 2014; Taspinar
et al., 2016)
1 (Fuß et al., 2014)
Meaning 1 (Taspinar et al., 2016) 1 (Taspinar et al., 2016) 1 (Taspinar et al., 2016) 1 (Taspinar et al., 2016)
Point 1 (Taspinar et al., 2016) 2 (Fuß et al., 2014; Taspinar et al., 2016) 2 (Fernandes and Junior, 2016; Taspinar et al., 2016)
Prize 1 (Fernandes and Junior, 2016) 1 (Fuß et al., 2014) 2 (AL-Smadi, 2015; Fernandes and Junior, 2016)
Reward schedule 1 (Fernandes and Junior, 2016) 1 (Fernandes and Junior, 2016)
Signposting 1 (AL-Smadi, 2015)
Social discovery 1 (Fuß et al., 2014)
Social network 2 (AL-Smadi, 2015; Harteveld and Sutherland,
2017)
Social pressure 2 (AL-Smadi, 2015; Harteveld and Sutherland, 2017)
Social status 2 (AL-Smadi, 2015; Fuß et al., 2014) 1 (Fuß et al., 2014)
Strategy 1 (AL-Smadi, 2015)
Time pressure 1 (Taspinar et al., 2016) 1 (Taspinar et al., 2016) 1 (Taspinar et al., 2016)
Virtual economy 1 (AL-Smadi, 2015)
Voting 1 (AL-Smadi, 2015)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
11
Six studies employed the Hexad typology (Marczewski, 2015). The
hexad typology describes six player types according to their motiva-
tions in gamified systems: achievers, disruptors, free spirits, philan-
thropists, players, and socializers. Achievers are intrinsically motivated
by competence and mastery, always trying to learn new things and
overcoming challenges. Disruptors are motivated by changes, either
negative (e.g., chasing others or discovering system crashes to spoil
others’ experiences) or positive ones (e.g., influencing others or ad-
justing the system flaws encountered). Free Spirits are intrinsically
motivated by autonomy and self-expression, enjoying exploring the
system with no restrictions and building new things. Philanthropists are
intrinsically motivated by meaning and purpose, being altruistic and
helping other players without expecting a reward for it. Players are
those extrinsically motivated by rewards, appreciating every tangible or
intangible benefit that the system may give them for their behavior.
Socializers are intrinsically motivated by relationships, enjoying to in-
teract with other users, and create social connections. Also, just like
Bartle typology, these player types are not mutually exclusive, and a
questionnaire aids in identifying the player type composition for each
user.
As shown in Table 7, there is a strong evidence in the literature that
achievers enjoy challenges and levels; free spirits relish unlockables
and customization; players like leaderboards and social status; and
socializers appreciate social networks and competition. Half of the
Hexad studies also suggest learning, social status, and unlockables to
achievers; customization to disruptors; exploration to free spirits;
gifting to philanthropists; badges, collections, competition, points,
prizes, unlockables, and virtual economy to players; and social status to
socializers. Challenges, customization, learning, and levels were game
elements suggested, in greater or lesser quantity, for all Hexad player
types.
There is also a divergence in the number of different game elements
explored for Hexad player types. Of the 32 distinct game elements
proposed for the Hexad typology, the literature suggested 25, 22 and 20
different game elements for achievers, players, and socializers (re-
spectively), while the studies that analyzed philanthropists, free spirits
and disruptors only proposed 18, 16 and 11 game elements. Still, when
observing the total of game elements proposed by each player type, we
have: 44 for players, 38 for achievers, 32 for socializers, 23 for free
spirits, 22 for philanthropists, and 21 for disruptors. The average il-
lustrates that free spirits had a centralized use of a more limited set
game elements, while the few elements studied for disruptors and
philanthropists were also little used. Similar to Bartle, it is not clear if
disruptors and philanthropists are less likely to enjoy the game ele-
ments or not, neither the influence of player types’ percentage in these
and future results.
The BrainHex typology relates the game motivation to the neu-
robiological reactions of the human body, defining seven different
player types: achievers, conquerors, daredevils, masterminds, seekers,
survivors, and socializers (Nacke et al., 2011; 2014). Achievers are those
motivated by overcoming challenging long-term goals and completing
collections since the satisfaction of success generates a fixation to
achieve new goals, supported by dopamine. Conquerors are those mo-
tivated by victories and the defeat of challenging enemies, since they
channel their anger to boost achievement, releasing norepinephrine and
testosterone into the body. Daredevils are players motivated by the thrill
of pursuit and risk-taking, releasing epinephrine (adrenaline) in the
body, which enhances the pleasure of conquest. Masterminds are players
motivated by the difficulty of solving problems and the strategy that
these solutions require, stimulating the production of dopamine, which
makes the achievement of goals intrinsically rewarding. Seekers are
those motivated by discovering new things or finding familiar things,
since their body produces endomorphin when visualizing images with
richly interpretable patterns, triggering their center of pleasure. Survi-
vors are players motivated by terror and the intensity of the associated
experience, releasing adrenaline into their body, which drives the
effects of dopamine when they hit their target. Socializers are players
motivated by interacting and trusting others since the primary neural
source activated is oxytocin, a neurotransmitter associated to trust.
These player types are not mutually exclusive, being discovered
through a questionnaire.
Of the four studies suggesting game elements for BrainHex player
types, three of them proposed the use of levels to achievers and lea-
derboards to conquerors, while two recommended time pressure to
achievers and daredevils, and signposting to socializers, as shown in
Table 8. No game element was suggested for all player types. In addi-
tion to exploring a smaller amount of game elements, these studies were
mostly examining different ones, given the large number of unitary
representations of game elements for each type of player. Achievers and
conquerors had the biggest number of game elements suggested (i.e.,
seven), and survivors were the player type less explored by BrainHex
studies. Again, we cannot affirm if the game elements preferred by
survivors were just less explored or if this player type is less likely to
enjoy gamification.
Three other studies used the typology proposed by Ferro et al.,
which is a correlation between many player typology proposals (Bartle,
1996; Bateman and Boon, 2005; Caillois, 2001; Fritz, 2004; Fullerton
et al., 2008; Koster, 2004; Nacke et al., 2011; Yee, 2006) and person-
ality traits studies (Berecz, 2008; Cattell, 2001; Crowne, 2007; Ryan
and Deci, 2000).Based on this correlation, the authors proposed five
different player types: creative, dominant, humanist, inquisitive, and
objectivist. Creative individuals like to create and develop things by
using skills obtained through experimentation. They enjoy having a
structured path, but with the possibility to treat it as a guide rather than
a directive. Dominant users exhibit a strong need to be visible, whether
through sociability, assertiveness, or aggressiveness. These individuals
prefer mechanics that are self-serving and personally related to their
participation. Humanists are more inclined to be social and involve
themselves in tasks that rely on social engagement. They like to work
with others to solve problems collaboratively rather than on their own.
Inquisitive users enjoy to explore and investigate new things. They are
more inclined to engage with open worlds, be in control, and embark on
quests to locate particular items. Objectivist is someone who seeks to
achieve and build upon their knowledge by demonstrating their ability
and intelligence. They are not necessarily as selfish as those who fall
into the Dominant category, but their focus is on their selves before
others. All three studies suggested collections and customization to
creative users, as Table 9 illustrates. Two studies also indicated the
preference of competition, customization, and leaderboards from
dominant users; consequences, emotions, and narrative from in-
quisitive ones; and challenges, levels, and meaning from objectivist
ones, while humanist users were the less examined Ferro et al. player
type. No game element was suggested for all player types proposed by
Ferro et al.. However, it is unclear how these three studies were able to
identify the player type of each user (e.g., questionnaire, interaction
data). Another interesting point is that while both Ferro et al. (2013)
and Monterrat et al. (2014, 2015b) were related educational contexts,
some divergences appeared. For instance, challenges were suggested to
creative, humanist, and inquisitive players by one study (Ferro et al.,
2013), and objective players by the others (Monterrat et al., 2014;
2015b). These divergences might indicate that one single user’s char-
acteristic may not be enough in promoting this tailoring effect.
The work of Barata et al. proposed four student profiles based on
their performance and gaming preferences, namely: achiever, regular,
halfhearted, and underachiever. Achievers were those focused on
reaching achievable goals and acquiring all possible points, excelling in
every aspect of the course, and being the most participative ones.
Regular students were those with performance above the average and
equilibrium between achievements and traditional evaluation compo-
nents. Halfhearted students were those who neglected some aspects of
the course and performed below the average. Underachievers were those
with the lowest performance, apparently making just the necessary
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
12
Table 7
Suggested game elements for each player type of Hexad typology.
Game element Achiever Disruptor Free spirit Philanthropist Player Socializer
Anarchy 1 (Tondello et al., 2016) 1 (Tondello et al., 2016)
Anonymity 1 (Tondello et al., 2016) 1 (Tondello et al., 2016) 1 (Tondello et al., 2016)
Badge 3 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2017a)
Challenge 5 (Challco et al., 2014; Herbert et al.,
2014; Holmes et al., 2015; Tondello
et al., 2017a; 2016)
2 (Tondello et al., 2017a;
2016)
2 (Tondello et al., 2017a; 2016) 1 (Tondello et al., 2017a) 2 (Tondello et al., 2017a; 2016) 1 (Tondello et al., 2016)
Choice 1 (Holmes et al., 2015) 1 (Tondello et al., 2016) 1 (Tondello et al., 2016)
Collection 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 2 (Holmes et al., 2015;
Tondello et al., 2017a)
3 (Holmes et al., 2015; Tondello
et al., 2017a; 2016)
Competition 1 (Tondello et al., 2017a) 2 (Orji et al., 2018; Tondello
et al., 2017a)
3 (Orji et al., 2018; Tondello
et al., 2017a; 2016)
4 (Holmes et al., 2015; Orji et al.,
2018; Tondello et al., 2017a; 2016)
Consequence 1 (Orji et al., 2018) 1 (Orji et al., 2018)
Customization 1 (Tondello et al., 2017a) 3 (Holmes et al., 2015; Orji
et al., 2018; Tondello et al.,
2016)
4 (Challco et al., 2014; Herbert et al.,
2014; Orji et al., 2018; Tondello
et al., 2016)
1 (Tondello et al., 2017a) 1 (Holmes et al., 2015) 2 (Orji et al., 2018; Tondello et al.,
2017a)
Easter egg 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 2 (Tondello et al., 2017a; 2016) 1 (Tondello et al., 2017a)
Exploration 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 3 (Holmes et al., 2015; Tondello
et al., 2017a; 2016)
1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Feedback 1 (Holmes et al., 2015) 1 (Orji et al., 2018)
Gifting 1 (Tondello et al., 2017a) 3 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2017a)
1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Guild 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Orji et al., 2018) 2 (Orji et al., 2018; Tondello et al.,
2016)
Leaderboard 1 (Tondello et al., 2017a) 4 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2017a; 2016)
1 (Tondello et al., 2017a)
Learning 3 (Holmes et al., 2015; Tondello
et al., 2017a; 2016)
1 (Tondello et al., 2017a) 2 (Tondello et al., 2017a; 2016) 1 (Tondello et al., 2016) 2 (Tondello et al., 2017a; 2016) 1 (Tondello et al., 2016)
Level 4 (Herbert et al., 2014; Holmes et al.,
2015; Tondello et al., 2017a; 2016)
1 (Tondello et al., 2016) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 2 (Tondello et al., 2017a; 2016) 1 (Tondello et al., 2017a)
Lottery 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Meaning 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 2 (Holmes et al., 2015;
Tondello et al., 2017a)
1 (Tondello et al., 2017a) 1 (Orji et al., 2018)
Narrative 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Point 1 (Tondello et al., 2017a) 3 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2016)
Prize 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 3 (Orji et al., 2018; Tondello
et al., 2017a; 2016)
2 (Orji et al., 2018; Tondello et al.,
2017a)
Signposting 1 (Tondello et al., 2017a)
Social discovery 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2016) 2 (Tondello et al., 2017a; 2016)
Social network 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 2 (Herbert et al., 2014; Tondello
et al., 2017a)
5 (Challco et al., 2014; Herbert et al.,
2014; Holmes et al., 2015; Tondello
et al., 2017a; 2016)
Social pressure 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Social status 3 (Herbert et al., 2014; Tondello
et al., 2017a; 2016)
1 (Herbert et al., 2014) 4 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2017a; 2016)
3 (Challco et al., 2014; Herbert et al.,
2014; Tondello et al., 2017a)
Strategy 1 (Holmes et al., 2015)
(continued on next page)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
13
Table 7 (continued)
Game element Achiever Disruptor Free spirit Philanthropist Player Socializer
Unlockable 3 (Challco et al., 2014; Herbert et al.,
2014; Tondello et al., 2017a)
1 (Tondello et al., 2017a) 5 (Challco et al., 2014; Herbert et al.,
2014; Holmes et al., 2015; Tondello
et al., 2017a; 2016)
1 (Tondello et al., 2017a) 3 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2017a)
Virtual economy 1 (Tondello et al., 2017a) 3 (Challco et al., 2014; Herbert
et al., 2014; Tondello et al.,
2016)
Voting 1 (Tondello et al., 2017a) 2 (Tondello et al., 2017a;
2016)
1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Table 8
Suggested game elements by each player type of BrainHex typology.
Game element Achiever Conqueror Daredevil Mastermind Seeker Socializer Survivor
Competition 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al.,
2014)
Consequence 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al., 2014)
Customization 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al., 2014)
Feedback 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al., 2014) 1 (Orji et al.,
2014)
Guild 1 (Orji et al., 2014) 1 (Orji et al., 2014)
Leaderboard 1 (Monterrat et al., 2015a) 3 (Lavoué et al., 2018; Monterrat
et al., 2015a; 2017)
1 (Monterrat et al., 2017)
Learning 1 (Monterrat et al., 2017) 1 (Monterrat et al., 2017)
Level 3 (Lavoué et al., 2018; Monterrat
et al., 2015a; 2017)
1 (Monterrat et al., 2015a) 1 (Monterrat et al.,
2015a)
Prize 1 (Orji et al., 2014)
Signposting 1 (Monterrat et al.,
2015a)
2 (Lavoué et al., 2018; Monterrat
et al., 2015a)
Social status 1 (Orji et al., 2014)
Time pressure 2 (Lavoué et al., 2018; Monterrat
et al., 2015a)
1 (Monterrat et al., 2015a) 2 (Lavoué et al., 2018; Monterrat
et al., 2015a)
Unlockable 1 (Lavoué et al., 2018) 1 (Lavoué et al., 2018)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
14
minimum effort to pass the course. Since it is still the only study about
this typology, Barata et al. (2016) suggest the use of challenges, points,
and badges to attract achievers, regular and halfhearted students,
while further proposing learning to achievers. No game elements were
suggested to underachievers. Similar to other typologies, we are still
unaware if these students were performing just the necessary to pass the
course because they are not motivated enough by the explored game
elements or if they are not driven by any game element at all.
The work of Borges et al. created five player roles based on three
player typologies (Bateman and Boon, 2005; Ferro et al., 2013; Yee,
2006): achievers, conquerors, creators, explorers, and humanists.
Achievers are players that enjoy winning and accumulating all available
rewards (defined as goal-oriented). Conquerors are those who ap-
preciate testing their skills and competing against others (people-or-
iented). Creators are players who like customizing the system (system-
oriented). Explorers enjoy inspecting the system by discovering its ins-
and-outs (also system-oriented). Humanists are those that enjoy so-
cializing, sharing learning, and relationship building (also people-or-
iented). The authors suggested badges, challenges, honor system, lea-
derboards, levels, points, and virtual economy to achievers; and
customization, learning, and levels to creators. Besides being the only
study about this typology, no game elements were proposed to con-
querors, explorers and humanists. Like the typology by Ferro et al.,
the methods to identify the player types of each user are not explicitly
defined.
4.3.2. Gender
The second most investigated characteristic was gender, explored by
seven studies. One study also examined the femininity/masculinity di-
mension of gender identity (Busch et al., 2016), which refers to “the
degree to which persons see themselves as masculine or feminine given
what it means to be a man or woman in society” (Stets and
Burke, 2000). Inside this dimension, femininity aspects include being
“affectionate”, “gentle”, and “sensitive to the needs of others”, while
masculinity includes “acting as a leader”, “making decisions easily”,
and “willing to take risks” (Stets and Burke, 2000).
Three out of these seven studies suggested badges, customization,
leaderboards, and levels to women, and competition to men. Two
studies also recommend guilds, points, and social status to women; and
customization, guilds, and prizes to men, as shown in Table 10. The
study of Busch et al. (2016) also advised the use of competition, con-
sequences, customization, feedback, prizes, signposting, and social
status to femininity. At the same time, no game elements were sug-
gested to the masculinity dimension of gender identity. Given that
only one study analyzed gender and the femininity/masculinity di-
mension of gender identity together, little is known about how much
one may influence in the other (or even the impact of any other gender
identity dimension over these two characteristics).
4.3.3. Personality traits
The third most explored characteristic were personality traits,
where five out of six studies based their findings on a taxonomy called
Big Five (also known as the five-factor model or the OCEAN model).
This model defines five fundamental factors to describe most person-
ality traits: agreeableness, conscientiousness, extroversion, neuroticism,
and openness to experience (Roccas et al., 2002). Agreeableness en-
compasses individual propensity to social harmony; Conscientiousness
measures the personal reliability based on aspects like organization and
responsibility; Extroversion indicates the comfort level of an individual
with relationships; Neuroticism represents the tendency of experiencing
negative emotions (e.g., sadness, anxiety, anger); and Openness to ex-
perience focus on the individuals’ range of interests with any novelty
(Rizvi and Fatima, 2015). For each factor, an individual can have dif-
ferent polarities. In essence, an altruistic person (i.e., high agreeable-
ness) could also have depression feelings (e.g., high neuroticism) in a
chaotic life (i.e., low conscientiousness). In contrast, another person can
Table 9
Suggested game elements by each player type of Ferro et al. typology.
Game element Creative Dominant Humanist Inquisitive Objectivist
Badge 1 (Ferro et al., 2013) 1 (Ferro et al., 2013)
Challenge 1 (Ferro et al., 2013) 1 (Ferro et al., 2013) 1 (Ferro et al., 2013) 2 (Monterrat et al., 2014; 2015b)
Collection 3 (Ferro et al., 2013; Monterrat et al., 2014; 2015b)
Competition 2 (Monterrat et al., 2014; 2015b)
Consequence 2 (Monterrat et al., 2014; 2015b)
Customization 3 (Ferro et al., 2013; Monterrat et al., 2014; 2015b) 2 (Monterrat et al., 2014; 2015b) 1 (Ferro et al., 2013)
Emotion 2 (Monterrat et al., 2014; 2015b)
Leaderboard 1 (Ferro et al., 2013)
Level 2 (Ferro et al., 2013; Monterrat et al., 2014) 2 (Ferro et al., 2013; Monterrat et al., 2014)
Meaning 2 (Monterrat et al., 2014; 2015b)
Narrative 2 (Monterrat et al., 2014; 2015b)
Point 1 (Ferro et al., 2013)
Prize 1 (Ferro et al., 2013) 1 (Ferro et al., 2013)
Reward schedule 1 (Ferro et al., 2013) 1 (Ferro et al., 2013)
Social status 1 (Ferro et al., 2013)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
15
be creative (i.e., high openness to experience) but also very shy (i.e.,
low extroversion). Each user has to complete a questionnaire to identify
the polarity of each factor.
Three studies suggested the use of customization and leaderboards
to individuals with high extroversion, and two of them recommended
badges, competition, feedback, levels, meaning, points, and social
networks to extroverted people; badges, levels, and prizes for people
with high neuroticism; and customization to individuals with high
openness to experience. As Table 11 illustrates, some of the polarities
have not been explored by the literature, like low agreeableness, con-
scientiousness, and neuroticism. By far, the most investigated factor
was extroversion, with a total of 32 game elements suggestions, which
represents 49% of all suggestions to Big Five taxonomy.
The other study that analyzed the effects of personality traits on the
game elements used the Myers-Briggs Type Indicator (MBTI), which
is a questionnaire grounded on Jung personality theory that describes
the individual’s experience according to four preferences: world, in-
formation, decision-making, and attitude (Myers et al., 1998). Like the
Big Five factors, these four preference factors also have polarities.
World preference corresponds to where an individual lives most in: ei-
ther inner (introverted) or outer (extroverted) – equivalent to the Ex-
troversion factor on the previous personality traits taxonomy. In-
formation preference refers to the depth of information desire: sensing
personalities focus on essential data and real facts, while intuition fo-
cuses on identifying patterns and adding meaning to it – related to
Openness to Experience factor. Decision-making preference identifies
how an individual makes choices: if he/she puts more weight on ob-
jective principles and facts (i.e., thinking) or on personal concerns and
people involved (i.e., feeling) – similar to the Agreeableness factor.
Attitude preference reports the way a person deals with the outside
world: preferring making decisions (i.e., judging) or staying open to
new options (i.e., perceiving) – linked to the Conscientiousness factor.
Although there is no equivalent to Neuroticism from the Big Five per-
sonality traits in this taxonomy, each individual is again a combination
of polarities for all four factors, which are also identified through a
questionnaire.
Based on this only study (Butler, 2014), the suggested game ele-
ments for the world preference factor are choices, customization, nar-
rative, signposting, and social networks to extroverts; and con-
sequences, meaning, and single-player for introverts. For the
information preference factor, the indicated game elements are chal-
lenges, collection, learning, levels, and meaning for sensing polarity;
and choices, exploration, fixed rewards, and narrative for intuitive
individuals. For the decision-making preference factor, the authors
suggested challenges, guilds, and strategy for feeling polarity; while
using badges, competition, levels, and social status for thinking po-
larity. For attitude preference, the authors suggested choices, collec-
tion, levels, strategy, and time pressure for judging polarity; and an-
archy, challenges, consequences, customization, easter eggs, and
emotions for perceiving polarity.
Compared to Big Five, both taxonomies considered employing cus-
tomization, signposting, and social networks to extroverts. None of the
game elements proposed to the Information preference factor was also
indicated for Openness to experience. Individuals with high agree-
ableness or feeling polarity would both appreciate challenges and
guilds. While levels were suggested to high conscientiousness and
judging individuals, the indication of consequences and customization
to perceiving polarity contradicts the Big Five analysis. Unlike in the
case of the Big Five model, all polarities of the four factors from MBTI
had at least three game element suggestions, while being suggested by
only one study.
4.3.4. Motivation
Four studies analyzed how user motivation correlates to game ele-
ment preferences. Three of them focused on the Goal Orientation
Theory, explored by Nicholls (1984) and Elliot (1999). This theory
defines how individuals interpret and experience achievement settings,
divided into two interrelated dimensions: mastery/performance goals,
and approach/avoidance motivation valence. Mastery goals focus on
developing competence (i.e., task mastery), whereas performance goals
focus on demonstrating it (i.e., bragging about it to others). Approach
motivation is related to a “behavior [that] is instigated or directed by a
Table 10
Suggested game elements by gender.
Gender Gender identity
Game element Women Men Femininity Masculinity
Badge 3 (Codish and Ravid, 2017; Denden et al., 2017b; Tondello et al.,
2017a)
Challenge 1 (Tondello et al., 2017a)
Choice 1 (Tondello et al., 2017a)
Collection 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Competition 3 (Busch et al., 2016; Oyibo et al., 2017a;
2017b)
1 (Busch et al., 2016)
Consequence 1 (Orji, 2014) 1 (Busch et al., 2016)
Customization 3 (Denden et al., 2017b; Orji, 2014; Tondello et al., 2017a) 2 (Denden et al., 2017b; Tondello et al., 2017a) 1 (Busch et al., 2016)
Feedback 1 (Denden et al., 2017b) 1 (Denden et al., 2017b) 1 (Busch et al., 2016)
Gifting 1 (Tondello et al., 2017a)
Guild 2 (Busch et al., 2016; Orji, 2014) 2 (Busch et al., 2016; Tondello et al., 2017a)
Leaderboard 3 (Codish and Ravid, 2017; Denden et al., 2017b; Tondello et al.,
2017a)
1 (Codish and Ravid, 2017)
Learning 1 (Tondello et al., 2017a)
Level 3 (Codish and Ravid, 2017; Denden et al., 2017b; Tondello et al.,
2017a)
1 (Denden et al., 2017b)
Lottery 1 (Tondello et al., 2017a)
Meaning 1 (Tondello et al., 2017a)
Point 2 (Codish and Ravid, 2017; Tondello et al., 2017a) 1 (Codish and Ravid, 2017)
Prize 1 (Tondello et al., 2017a) 2 (Oyibo et al., 2017a; 2017b) 1 (Busch et al., 2016)
Signposting 1 (Tondello et al., 2017a) 1 (Busch et al., 2016)
Social discovery 1 (Tondello et al., 2017a)
Social network 1 (Tondello et al., 2017a)
Social status 2 (Orji, 2014; Tondello et al., 2017a) 1 (Busch et al., 2016)
Virtual economy 1 (Tondello et al., 2017a)
Voting 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
16
Table 11
Suggested game elements by each personality trait of Big Five taxonomy.
Agreeableness Conscientiousness Extroversion Neuroticism Openness
Game element High Low High Low High Low High Low High Low
Badge 2 (Codish and Ravid, 2014; Denden et al.,
2017a)
1 (Denden et al.,
2017a)
2 (Jia et al., 2016; Tondello
et al., 2017a)
Challenge 1 (Jia et al., 2016) 1 (Tondello et al., 2017a)
Collection 1 (Tondello et al., 2017a)
Competition 1 (Orji et al.,
2017)
2 (Orji et al., 2017; Tondello et al.,
2017a)
1 (Orji et al.,
2017)
Consequence 1 (Orji et al.,
2017)
1 (Orji et al., 2017) 1 (Orji et al., 2017) 1 (Orji et al.,
2017)
Customization 1 (Orji et al.,
2017)
1 (Orji et al., 2017) 3 (Denden et al., 2017a; Orji et al., 2017;
Tondello et al., 2017a)
1 (Denden et al.,
2017a)
2 (Orji et al., 2017; Tondello
et al., 2017a)
1 (Jia et al., 2016)
Feedback 1 (Orji et al.,
2017)
1 (Orji et al., 2017) 2 (Denden et al., 2017a; Orji et al., 2017) 1 (Denden et al.,
2017a)
1 (Orji et al.,
2017)
Guild 1 (Orji et al.,
2017)
1 (Orji et al., 2017) 1 (Orji et al.,
2017)
Leaderboard 3 (Denden et al., 2017a; Jia et al., 2016;
Tondello et al., 2017a)
Level 1 (Jia et al., 2016) 2 (Denden et al., 2017a; Jia et al., 2016) 2 (Jia et al., 2016; Tondello
et al., 2017a)
Lottery 1 (Tondello et al., 2017a)
Meaning 1 (Orji et al., 2017) 2 (Orji et al., 2017; Tondello et al.,
2017a)
Point 2 (Denden et al., 2017a; Jia et al., 2016) 1 (Jia et al., 2016) 1 (Tondello et al., 2017a)
Prize 1 (Orji et al.,
2017)
1 (Orji et al., 2017) 2 (Jia et al., 2016; Tondello
et al., 2017a)
1 (Tondello et al., 2017a) 1 (Orji et al.,
2017)
Signposting 1 (Tondello et al., 2017a)
Social network 2 (Denden et al., 2017a; Tondello et al.,
2017a)
1 (Denden et al.,
2017a)
Social pressure 1 (Tondello et al., 2017a)
Social status 1 (Tondello et al., 2017a) 1 (Tondello et al., 2017a)
Virtual economy 1 (Tondello et al., 2017a)
Voting 1 (Tondello et al., 2017a)
A.C.T. Klock, et al. International Journal of Human-Computer Studies 144 (2020) 102495
17
positive or desirable event or possibility” Elliot, while negative or un-
desirable events and possibilities direct the avoidance motivation be-
haviors. Thus, this theory presents four combinations for these dimen-
sions: Mastery-Approach, where users tend to improve skills; Mastery-
avoidance, where users avoid facing incompetence; Performance-ap-
proach, where users tend to demonstrate their skills to others; and
Performance-avoidance, where users avoid demonstrating incompetence.
A questionnaire aids in identifying the goal orientation of each in-
dividual.
The literature explored only three game elements: badges, feedback,
and leaderboards, but only one of these elements was investigated by
more than one study. All of the works suggested badges to users in the
performance-approach spectrum and two of these studies also in-
dicate the use of badges to mastery-approach users, as shown in
Table 12. Badges are suggested to almost every combination except
performance-avoidance, whose preference relies on leaderboards ac-
cording to one single study.
The other study analyzed the Murray’s secondary psychological
needs. Besides the primary needs (e.g., food, water, air), Murray
grouped secondary needs by domains: affection, ambition, information,
materialism, and power, these being related to the eight suggested
game elements. Affection is related to the desire to love and be loved;
Ambition is associated with the need for achievement and recognition;
Information aims to gain knowledge and share it with others; Materi-
alism relates to acquisition, construction, order, and retention of ma-
terial goods; and Power relies on independence and on the need to
control others (Flett, 2007). The suggested game elements are compe-
titions and guilds for affection; badges, challenges, feedback, and
meaning for ambition; feedback and learning for information; custo-
mization for materialism; and challenges and competition for power
(Berg and Petersen, 2013). The most used game elements were chal-
lenges, competition, and feedback (suggested for two different needs),
while the need with more indicated game elements was ambition (four
game elements). Still, only one study supports these results.
4.3.5. Age
Three studies also based their analysis on the age of the users, which
states the most suitable game elements for younger people (i.e., 18–25
years old in Oyibo et al. (2017a), less than 25 years old in
Oyibo et al. (2017b), and 30 years old or less in
Tondello et al. (2017a)). With a total of 21 different game elements
explored, only two of them were suggested in more than one study:
competition and learning (Oyibo et al., 2017a; 2017b). The other game
elements considered exclusively by Tondello et al. (2017a) were:
badges, challenges, collection, customization, exploration, gifting,
guilds, levels, lottery, meaning, points, prizes, social discovery, social
pressure, social networks, social status, unlockables, virtual economy,
and voting. No game element was indicated simultaneously by all three
works, and there were no suggestions for users who are more than 30
years old.
4.3.6. Behavior
Three studies investigated the effects of game elements on user
behavior (Denden et al., 2017b; Utomo and Santoso, 2015; Xu and
Tang, 2015). By behavior, two of works mean the access frequency:
while Utomo and Santoso (2015) analyzed the exact average period
between accesses (e.g., every three days, every two weeks),
Denden et al. (2017b) divided users in two groups: regular ones (i.e.,
daily or weekly access) and non-regular ones (i.e., monthly access or a
more spaced frequency). Badges were the only game element explored
by the two studies, being suggested to every user that accesses the
system at least every two weeks, and to regular and non-regular users.
All other six game elements were proposed only by one study: time
pressure for every user that accesses the system at least every three days
(Utomo and Santoso, 2015); customization, feedback, leaderboards,
and levels for regular users; and points for both, regular and non-reg-
ular, ones (Denden et al., 2017b). Meanwhile, Xu and Tang (2015)
employed data mining techniques in users’ action to establish a gami-
fication recommendation system but did not specify the explored game
elements.
4.3.7. Culture
Culture is here considered as “the collective programming of the
mind that distinguishes the members of one group or category of people
from others” (Hofstede, 1984), divided in five dimensions: Femininity
versus Masculinity (also called gender identity, described previously on
Subsection 4.3.2 in conjunction with gender), Individualism versus
Collectivism, Long-term versus Short-term orientation, Power Distance
(i.e., power distribution within society), and Uncertainty Avoidance
(i.e., tolerance to unpredictability) (Hofstede et al., 2010). The two
studies described here focused on Individualism versus Collectivism,
which divides scenarios where people are expected to look after
themselves and their significant others first (i.e., individualists), and
where people are expected to look after and protect other in-group
members in exchange for unquestioning loyalty (i.e., collectivists)
(Oyibo et al., 2017a; 2017c).
The explored game elements by these two works were: competition,
learning, and prizes. While prizes were suggested to collectivist in-
dividuals by the two studies; competition was a collectivism preference
in