Access to this full-text is provided by IGI Global Scientific Publishing.
Content available from International Journal of Mobile and Blended Learning
This content is subject to copyright. Terms and conditions apply.
DOI: 10.4018/IJMBL.318223
Volume 15 • Issue 2
This article published as an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0/) which permits unrestricted use, distribution, and production in any medium,
provided the author of the original work and original publication source are properly credited.
*Corresponding Author
1
Daniel Biedermann, DIPF Leibniz Institute for Research and Information in Education, Germany
Patrick Oliver Schwarz, DIPF Leibniz Institute for Research and Information in Education, Germany
Jane Yau, DIPF Leibniz Institute for Research and Information in Education, Germany*
Hendrik Drachsler, DIPF Leibniz Institute for Research and Information in Education, Germany
App-based habit building has been shown to be a good tool for forming desired habits; however, it is
unclear how much individual features that are present in many apps contribute to the success of habit
building. In this paper, the authors consider the influence of social support features by developing an
app in which habit progress was shared with peers – ‘buddies’ in the app. In the study, 38 participants
created habits and monitored their progress regularly with the app over three weeks. The participants
were divided into a control group without a ‘buddy’ and a treatment group cohort in which they were
assigned to buddies based on their desired habits. With each habit repetition, the app gave feedback
on the number of repetitions and the automaticity of the user’s habit. The results obtained show that
the reproduction of app-based intentional habit building is effective and that automaticity could be
predicted by habit repetition.
Habit Forming, Self-Determination Theory, Social Support Features
There is a considerable number of people who struggle with the technique of forming habits that could
improve their learning processes, thereby making them more effective and efficient. Especially in
the context of technology-enhanced learning, and due to the increased number of technology-related
distractions (e.g., advertisements, temptations to browse other websites), the act of forming desirable
or undesirable habits significantly influences the learning process and study success (Fiorella, 2020).
Habits can influence learning in a positive way by ensuring regular and consistent learning efforts.
At the same time, habits can also have a negative impact, for example when habitual excessive media
consumption leads to continual distractions (Lee, 2014). Habits are behavioral patterns that are
triggered by a particular context, often outside of conscious awareness (Pinder & Cowan, 2018). A
student might have the habit of regularly checking their mobile phone, even when studying. However,
Volume 15 • Issue 2
2
even though habits are mostly triggered automatically, they can still be actively, consciously, and
intentionally learned. Lally et al. (2010) showed in their study that participants were able to build
up habits over a period of several weeks in their natural environment. They found that the growth
of habit strength, which they called automaticity, can be described by a quadratic function that first
increases sharply and then stagnates once the asymptote / tangent is reached. This means that the
initial repetitions cause a high growth of automaticity, which then decreases with each repetition until
the behavior reaches its limit of automaticity.
The possibility of helping users to form habits has been taken advantage of by the designers and
developers of over 100,000 mobile apps in recent years, particularly for health reasons such as exercise,
diet, and weight management (Edwards et al., 2016). With the ubiquitous use of smartphones today,
hundreds of millions of people use such apps for improving their lifestyle, health, study, and work
successes etc. (Ibid). A number of individual features are used by these apps, such as paying the user
a (virtual) reward for completing the target activity, or providing accurate feedback about the user’s
progress and performance towards the goal. In many situations, a social feature component (such as
sharing the progress with family or friends) has been found particularly effective for individuals to
achieve their goals (Villalobos-Zuniga & Cherubini, 2020). Additionally, features utilized by such
apps have been designed based on behavioral theories that focus on observable behavior, such as Self-
Regulation Theory (Bandura, 1986), Social Cognitive Theory (Bandura, 1986), Theory of Planned
Behaviour (Ajzen, 1985), Trans-Theoretical Model (Prochaska & Di Clemente, 1983), Health Belief
Model (Rosenstock, 1974), and Goal-Setting Theory (Locke and Latham, 2002). These theories are
typically used to explain the reasons for people undertaking (or not undertaking) a certain activity
and the different stages of progressing through it. Villalobos-Zuniga and Cherubini (2020) identified
a major common role / indicator, namely a person’s motivation for doing the task, as a decisive factor
for whether the task will be completed or not. They selected the Self-Determination Theory (SDT)
(Deci and Ryan, 2008) as the foundation of app features upon which their taxonomy was built, which
relates to different aspects of motivation. Broadly speaking, SDT can be classified into intrinsic /
internal motivation (e.g., studying for one’s own interests) and extrinsic / external motivation (e.g.,
studying because my family wants me to, or to get a good job). Note that we are often both intrinsically
and extrinsically motivated to carry out different tasks. Recently, many apps have utilized additional
internal or external incentives to help users to succeed more with completing an activity or reaching
a goal or making an activity becoming habitual. In spite of the prevalence of these apps and their
large number of users there is a lack of professional guidelines for designers, or industry standards,
and lacking knowledge on the long-term effects of such interventions means there are concerns that
such apps could even lead people to adopt the opposite of the target behavior, in the worst scenarios
(Edwards et al., 2016).
Building desirable habits, and getting rid of undesirable ones, has also become a major topic in
the digital behavior change literature (Pinder & Cowan, 2018). Habit building via self-monitoring
on smartphone apps over longer durations has been successfully demonstrated by Stojanovic et
al. (2020). The self-monitoring itself is only one feature of digital behavior change apps, and they
often utilize a wide range of other motivational features, such as reminders, gamification, or social
support (Villalobos-Zuniga & Cherubini, 2020). Despite these features being used often, there is a
lack of research regarding the contribution of these individual and/or social features for habit building
(Hermsen et al., 2016). Especially for social features, which are central to many (commercially
successful) apps, research is still in its infancy (Elaheebocus et al., 2018; Oinas-Kukkonen et al.,
2009). With this study, the authors aim to address this gap by exploring and examining the impact
of social support features in app-based habit building. For this investigation, the authors created an
app called Habit Buddy that allows habit creation and self-monitoring with additional social support
features. Users of the app can have a peer (henceforth called buddy) with whom they can communicate
and share their habit tracking progress, which is stored and analyzed. The authors then investigated
Volume 15 • Issue 2
3
the effect of these social support features. The objective of this research is guided by the following
research question:
1. Does habit automaticity increase with habit repetitions?
a. Is there a correlation between social support features and higher habit automaticity?
b. Is the effect of habit repetitions on automaticity stronger when accompanied by social support
features?
In this section, the authors provide a background literature review on habit forming and its relationship
with two different theories – Self-Regulation and Self-Determination (Bandura, 1986) - different
persuasive app features identified by Villalobos-Zuniga and Cherubini (2020), and the extension of
the authors’ previous work on digital self-control tools (Biedermann et al., 2021).
The use of habits can be a way for a learner to increase their self-regulation skills in relation to
undergoing and continuing with their learning processes and tasks. In a study by Breitwieser et al.
(2022), it was found that psychological interventions did not typically function as a one-off event,
but rather when these become repeated or regular motivational prompts, one can expect a higher
rate of achievement of learning goals. Habit forming can be seen as a way of helping a person to
become more self-regulated and self-directed within the context of the activity (e.g., to exercise, keep
a diary), if this activity is something that the person intentionally wants to form a habit of (Bailey et
al., 2020). In this sense, prompts to support repetition in the habit-forming process are more likely
to reinforce the behavior. There have been several studies showing that self-regulation skills are a
strong predictor of achievement, and therefore Breitwieser et al.’s (2022) experiment aimed to show
that repeated interventions, similar to those found in habit forming, can improve students’ use of
self-regulated learning strategies to achieve their goals.
In a study by Villalobos-Zuniga and Cherubini (2020), 208 apps were analyzed, which were identified
to contain 12 design features that could support users with behavioral adjustment (e.g., exercise more,
meditate more, quit smoking, lose weight). The authors classified these design features according
to the Self-Determination Theory. They identified a number of research gaps including whether the
use of such apps a) actually nurture or thwart intrinsic motivation, b) provide support to the three
basic needs – growth, well-being and integrity, and c) provide optimal challenge. Cooperation has the
possibility to influence habit forming by affecting each individual’s intrinsic motivation by focusing
and aiming to achieve the same/similar goals together. In particular, one app feature showed the
support of close contacts in achieving goals and this led to the individuals feeling a greater relatedness
and sense of belonging, which helped them sustain their self-determined motivation. The taxonomy
was constructed to help relevant stakeholders such as researchers to test and evaluate their app and
includes features / interventions or how a combination of these can be utilized to motivate users to
achieve their habit goals (Ibid).
Figure 1 below shows the process by which Villalobos-Zuniga and Cherubini (2020) created their
taxonomy from the reviewed apps and interventions, which were then categorized into three sets
of categories with different persuasive features. The first set is the Autonomy category, which
Volume 15 • Issue 2
4
consists of Reminders, Goal Setting, Motivational Messages, and Pre-commitment. The second set
is Competence, which consists of Activity Feedback, History, Log/Self-Monitoring, and Rewards.
The third set is Relatedness, which consists of Performance Sharing, Peers Comparison, Challenge
Peer, and Messaging.
Habits play a big role in how learners deal with digital distractions. Beneficial habits, such as turning
off the smartphone before starting to learn, can help reduce distractions and improve learning outcomes
(Galla & Duckworth, 2015). Building of “good” habits also helps in the interplay with other tools.
Common tools for reducing digital distractions include website blockers or visualizations of one’s
own behavior (c.f. Biedermann et al., 2021). Habits often prevent these tools from having their desired
effect. A visualization of one’s own usage behavior can only work if it leads to better habits. A website
blocker does not work if a user gets into the habit of always skipping it. When looking at the success
factors of such tools, one finds that intrinsic motivation to use them is a critical success factor.
Extending from this previous work, the research being addressed in this paper examines the
perspective of forming habits so that this becomes an intrinsic / internal motivated habit, whether
consciously or unconsciously. Thus, the authors’ research in this study is focused on whether social
support features could enhance habit forming potential and automaticity. In particular, the authors’
app has been developed mostly with the Relatedness category in mind because the authors were most
interested in the effects of peer and social influence on how habits could be potentially formed and
strengthened (Villalobos-Zuniga & Cherubini, 2020). This is also directly related to two theories /
Figure 1. Taken directly and permission given from Villalobos-Zuniga and Cherubini’s (2020) taxonomy creation process
Volume 15 • Issue 2
5
concepts – 1) Social Cognitive Theory (Bandura, 1986) which states that people often copy / replicate
the behavior of others as a social phenomenon, and peers could benefit from motivating one another
to achieve a common goal, and 2) the need to belong, which increases motivation (Baumeister &
Leary, 1995). The authors are specifically interested in how feedback and self-monitoring (individual
feature) as well as sharing of this progress with a peer (social feature) could help form habits more
effectively, and how intrinsic and extrinsic motivation can be increased as a result of the individual
and social features. From these app features for influencing the processes of forming beneficial
habits, and the research gaps mentioned above, the authors identified the potential of an app-based
habit formation approach in which users can cooperate and positively influence each other to work
toward achieving their goal.
The authors’ study was conducted in a between-subjects design with the dependent variable habit
automaticity, using an app with and without social support features for the participants to use in an
experimental setting. In order to prove that the app would be suitable for habit building in general, the
authors aimed to replicate the results central to app-based habit building in general, thereby proving or
disproving the hypothesis H1a. Thereafter, the authors examine whether the use of the additional social
support features correlates with higher habit automaticity (hypothesis H1b), and whether the effect of habit
repetitions on automaticity is stronger when accompanied by social support structures (hypothesis H1c).
H1a: Habit automaticity increases with habit repetitions.
H1b: The enabling of social support features leads to higher habit automaticity.
H1c: The effect of habit repetitions on automaticity is stronger when accompanied by social support
features.
The authors conducted a three-week experiment with 38 participants, where half of the participants
used a variant of the app with the social support features being enabled, and the other half used
the app without these features. The study participants were required to be at least 18 years old in
order to give their own consent for taking part. They were recruited through open calls via emails to
university students as well as via social media networks. Additionally, there was an incentive given
to participants in the advertisement that a fixed number of vouchers would be raffled among the
participants. N = 38 people (21 female, 17 male) participated, and they were divided into a Buddy
cohort and a Non-Buddy cohort. It was important that participants in the buddy cohort did not know
each other personally, as this would influence the effects of social interaction within the app. For this
reason, the group division resulted in a Buddy cohort of N0 = 18 subjects and a Non-Buddy cohort of
N1 = 20 subjects. Prior to the experiment, all participants were asked to fill in two questionnaires (1)
Self-Report Habit Index (SRHI) (Verplanken & Orbell, 2003) with the eight most relevant items for
the authors’ study (1, 2, 3, 4, 5, 6, 8, 9 from catalogue 4) in order to measure their habit strength and
automaticity, and a Self-Regulation Questionnaire (SRQ) by Brown et al. (1999). Each time that a
participant completed a habit repetition, he/she was required to answer two randomly selected items
from the SRHI in order to provide us with more insights into how or if their habits were forming as
a result of the habit repetition intervention on the mobile app. The SRHI was used because currently
there are no other tools available to assess more broadly how individuals form habits in daily life
and for seeking their own perception of this process (Ersche et al., 2017). The tool has also been
frequently utilized and highly cited by researchers indicating its usefulness and impact. Note that
the questions from the SRHI are built within the app on the habit repetition screen for the users to
answer. The SRQ was distributed prior to the start of the study.
Volume 15 • Issue 2
6
The app was implemented using the Flutter framework so that it would be available on both
iOS and Android platforms. The app allowed free creation and selection of new / existing habits, but
required those to be in specific categories such as the desire to be healthier, to be more organized, to
improve certain skills, to exercise more, to drink more water, or to improve one’s character (Figure 2a).
The purpose of this was so that participants with similar aspirations could be matched to each other
without having the exact same habit (e.g., placing the smartphone in a different room and turning the
smartphone app off could both be placed in the same category). The authors did not enforce specific
habits because the authors wanted to make sure that the participants picked something that they
truly cared about. In the authors’ app, the authors provided several categories and subcategories so
that every buddy pair in the authors’ experiment would have a larger selection of aspirations to work
from and in order to match them with others. In the habit creation screen (Figure 2a), participants
could give a name and a description to their aspiration that they wanted to form a habit routine of,
and decide whether they wanted to set a reminder in case they forgot this task. A small text box at the
bottom of the screen on the app appeared to remind the user of their habit-forming repetition routine
(Elaheebocus et al., 2018). Participants’ active habits were listed with an overview of the status of
the habit-forming routine, including the habit strength and whether the action was already performed
on that day (Figure 2b). In the app variant with social support features, the buddy’s motivation was
shown next to the habit repetition indicator. The motivation level was translated into a message if a
buddy had aspirations to form a habit of the same category. In Figure 2b, the topmost habit (Home
workout) is shared with a buddy. In the Habit Buddy-view (Figure 2c), the user can send messages
to their buddy (Figure 2b). They can furthermore view the history of their buddy’s habit repetitions.
Participants received weekly reminders in the three-week experiment to remind them to continue
using the app. They also received instructions via email, which were different depending on whether
they were in the Buddy or Non-Buddy cohort. Instructions to both cohorts differed in the section
where the function of the buddy was explained, which was missing for the Non-Buddy cohort. All
Figure 2. Screenshots that show screens of the Habit Buddy app (the authors’ own adaptation)
Volume 15 • Issue 2
7
participants were instructed to read the instructions carefully and were invited to ask questions at any
time, if anything was unclear. The instructions for habit creation stated that these categories should
be named or labelled as meaningfully as possible. There was no limit to the number of times that a
habit could be repeated in a day, with the aim of achieving automaticity.
In the course of the study, six of the participants became inactive and were subsequently removed
from the analysis. Since three female and three male participants were removed, this ratio did not
change. Note that the gender distribution is not used in the authors’ data analysis, but the authors
collected the gender information just as a reassurance to show that the gender distribution in the study
would be similar to a real-life situation. The 32 participants created a total of N = 93 habits (N0 =
44, N1 = 49). The average number of habits created per participant was M = 2.45 (M0 = 2.44 and M1
= 2.45). The habits were separated into “Eating healthier” (11.82%), “Organize better” (16.13%),
“Improve ability” (12.9%), “More exercise” (for example, take the stairs instead of the lift, or doing
more sports, with 23.66%), “Drink more water” (22.58%) and “Work on yourself” (For example,
discipline, hygiene or emotions, with 12.9%). The 32 participants generated a total of N = 609 habit
repetitions with an average of M = 22.71 (SD = 26.92) habit repetitions. There were 15 participants
in the Habit Buddy cohort with a total of N0 = 288 habit repetitions with an average of M0 = 19.20
(SD0 = 8.02). There were 17 participants in the Non-Buddy cohort with a total of N1 = 321 habit
repetitions and an average of M12 = 17.2 (SD12 = 7.74). Note that five participants continued to use
the app even after the end of the study.
H1a: Habit automaticity increases with habit repetitions.
H1b: The enabling of social support features leads to higher habit automaticity.
H1c: The effect of habit repetitions on automaticity is stronger when accompanied by social support
features.
In order to prove or disprove these hypotheses, all growth curves for the dependent variable
automaticity were examined with maximum likelihood parameter estimation. For each of the three
hypotheses, three different models were applied using random intercepts (𝜏0) and random slopes
(𝜏1). In all three models, time was also represented as the number of repetitions, similar to the work
in Stojanovic et al. (2020). For instance, at time t=10, a participant has completed a habit repetition
for the tenth time. All statistical calculations were carried out in Python and R.
Model 1 (H1a) consisted of the intercept-only model and was extended by the fixed-effect
predictor habit repetitions to a model that could predict the automaticity of a participant p at a time
t (Equation 1). The random intercept 𝛽0,𝑝 (Equation 2) consisted of the average intercept 𝛾0,0 and the
individual deviation 𝜏0,𝑝 from that intercept. The average slope 𝛽1,𝑝 (Equation 3), which is the slope
for the moderation effect of habit repetition, is given by 𝛾1,0 and 𝜏1,𝑝. Here, 𝛾1,0 is the average slope
of the whole sample and 𝜏1,𝑝 is the deviation from that slope.
Automaticity𝑡, 𝑝 = 𝛽0,𝑝 +𝛽1,𝑝 Habit Repetitions + 𝜀𝑡, 𝑝 (1) 𝛽0,𝑝 = 𝛾0,0 +𝜏0,𝑝 (2) 𝛽1,𝑝 = 𝛾1,0 +𝜏1,𝑝 (3)
For Model 2 (H1b, Equation 4), the presence of a buddy was added as a level 2 fixed effect —
(Equation 5). In this context, 𝛾0,1 is the effect of a buddy and is multiplied by the presence of a buddy.
If a user had a Habit Buddy, this was coded with a 1. If there was no buddy, this was coded with a 0.
The other parameters in this regression remain the same as in Equation 1.
Automaticity𝑡, 𝑝 = 𝛽0,𝑝 +𝛽1,𝑝 Habit Repetitions+ 𝜀𝑡, 𝑝 (4) 𝛽0,𝑝 = 𝛾0,0 +𝛾0,1 Has buddy1,𝑝+𝜏0,𝑝 (5)
𝛽1,𝑝= 𝛾1,0 +𝜏1,𝑝 (6)
Volume 15 • Issue 2
8
In Model 3 (H1c), a cross-level interaction was added (Equation 7). For the slope 𝛽1,𝑝 the effect of a
Habit Buddy was taken into account. This made it possible to investigate whether there were relationships
between the two predictors at level 1 and level 2. This was implemented by adding to 𝛾1,0 and 𝜏1,𝑝 the
moderation effect of a Habit Buddy multiplied by the presence of a Habit Buddy (Equation 9).
Automaticity𝑡, 𝑝 = 𝛽0,𝑝 +𝛽1,𝑝 Habit Repetitions + 𝜀𝑡, 𝑝 (7) 𝛽0,𝑝 = 𝛾0,0 +𝛾0,1 Has buddy1,𝑝 +𝜏0,𝑝 (8)
𝛽1,𝑝 = 𝛾1,0 +𝛾1,1 Has buddy1,𝑝 +𝜏1,𝑝 (9)
An overview in the form of a table is shown in Figure 3 below. In order to check the distribution across all
milestones, a Shapiro-Wilk test was applied, which suggested a left-skewed normal distribution (t(620)
= 0.969, p = 0.392). Based on the normal distribution assumption, a one-sample Welch’s t-test on both
milestone datasets of the Buddy and Non-Buddy cohort showed that the milestones of the Non-Buddy
cohort had the same mean as the mean of habit repetitions ((t1(19) = -2.758, p = 0.015), (t0(17) = -1.7,
p = 0.112)). In order to investigate the hypotheses H1a, H1b and H1c, the nature of the automaticity
datasets was examined. To ensure that there was actual space for automaticity to grow, only habits
with at least five repetitions were considered (N = 58). The tests showed that the automaticity values
per habit of the Non-Buddy group (t1(29) = 0.971, p = 0.557) and the Buddy group (t0(27) = 0.953,
p = 0.242) were normally distributed. However, the distribution of the Non-Buddy group was skewed
to the right and the Buddy group to the left. An unpaired Welch’s t-test showed that both samples have
identical means (t(56) = 0.296, p = 0.769). The same result occurred when comparing the mean values
with that of the total quantity (t0(84) = -0.218, p = 0.829), (t1(86) = 0.200, p = 0.843). An Intraclass
Correlation Coefficient analysis showed that over 47% of the variance of the automaticity data could be
explained by the person level (level 2) variance. This suggests that sufficient clustering takes place in
the inter-individual datasets to make the use of multilevel modelling of the data meaningful. To examine
whether the number of habit repetitions would predict automaticity (H1a, Figure 3), a model was built
with random intercepts. Habit repetition was then added as a fixed effect and random slopes were used
to ensure variation in the habit repetitions to make the model even more flexible. The result was that
habit repetition significantly predicted automaticity with a 𝛽1= 0.027 (SE = 0.028) and t(24.7) =4.163,
p = 0.000332. Furthermore, the intercept showed a significant variance (𝜏00= 0.024), which suggests
that the initial value of automaticity sometimes differed strongly within the users.
Figure 3. Multilevel regressions of automaticity on habit repetition & has-habit-buddy attribute (the authors’ own adaptation)
Volume 15 • Issue 2
9
In hypothesis H1b (Model 2, Figure 3), the existence of a Habit Buddy was added as a fixed
parameter to the model. The aim was to find out whether a Habit Buddy has a positive effect on
his partner. The extension of the model showed that having a Habit Buddy positively predicted the
automaticity with a 𝛽0 = 0.0441 (SE = 0.053). This value was over 64.5% higher than 𝛽1 nevertheless,
the habit buddy parameter had a very high p-value (t(35.14) = 0.833, p = 0.4103), which opens up the
likelihood that the predicted value of the automaticity might have been just as high without the Habit
Buddy. The habit repetition was significant as well in this model (t(24.74) = 4.166, p = 0.00329).
The effects of random effects remained almost unchanged.
To investigate hypothesis H1c, Model 2 was extended to include a cross-level interaction. By
adding this fixed effect, a level 2 predictor was also taken into account for the slope of the regression
line. With the resulting Model 3 (Model 3, Figure 3), it was thus possible to examine whether the
presence of a Habit Buddy had a different effect on the slope of the level 1 predictor. This cross-level
interaction made the model much more flexible in dealing with the hierarchical data. Again, habit
repetition significantly predicted automaticity (t(24.35) = 3.54, p = 0.00164). Once again, habit
repetition and having a habit buddy predicted automaticity, and both betas actually increased (𝛽1 =
0.0314, SE = 0.00887 and (𝛽0 = 0.0572, SE = 0.0557). However, only habit repetition continued
to show significance (t(24.353) = 3.54, p = 0.00164). It is interesting that the p-value of the Habit
Buddy parameter fell sharply and the t-value increased to over 1 (t(33.38) = 1.027, p = 0.31166).
Unfortunately, the direct cross-level effect was very small, regardless of whether a habit buddy was
present, the influence on the slope of the regression line was very negligible (Habit Repetitions *
Has Habit Buddy = -0.010, SE = 0.013). A t-test of this interaction yielded a t-value below zero and
a very high p-value (t(24.78) = 3.54, p = 0.454). Thus, as in Model 2, these values could also have
arisen by chance.
To determine the correlation between the participant’s self-regulation trait and the number of
habit repetitions, the Pearson product-moment correlation coefficient was used. The correlation was
low and not significant (r = 0.069, t (30) = 0.383, p = 0.704). A possible explanation for this was
that the participants were motivated to take part in this study, which might have caused the saturation
effect on habit automaticity by the high number of repetitions, and therefore leveling the effect of
the social support features.
CONCLUSION
Situations such as the COVID-19 pandemic have highlighted the need for students to have adequate
self-regulation (Daumiller & Dresel, 2019). Therefore, the forming of good habits and eliminating
bad ones can make one more resilient to distractions and help to achieve one’s goal. In this paper,
the authors presented the motivation, research question, hypotheses and data analysis to the research
problem relating to whether social support features in an app could help users in forming beneficial
habits. In both participant groups, the app demonstrated that it was possible to form habits using a
habit-forming mobile app, and habit automaticity increased predictably with habit repetitions. While
the habit building in general worked, the addition of social support (features) through having a Habit
Buddy did not lead to any additional improvements. The number of completed habit repetitions and
the automaticity of the participants showed no significant difference between the groups, where one
had social support features enabled in their app, and the other that did not have this feature. Although
there was a trend that indicated that a Habit Buddy had a positive influence on app-based habit
building, the results obtained in this study were not significant.
Potential reasons (and limitations for this research) include (1) participants were assigned
their buddies whom were not their mutually-close or trusted friend. In reality, the effect of habits-
forming may strengthen by carrying out this process together with a close friend with the same /
similar aspirations. This could possibly have led to the fact that they may not have been keen doing
this process with a stranger; (2) there was a ceiling effect in the sense that participants had already
formed their habit through a sufficient number of habit repetitions and therefore did not repeat this
behavior any more, or, conversely (3) due to the time constraints of the three-week experiment, some
Volume 15 • Issue 2
10
habits may not have had enough time to be formed and developed at all, (4) the experiment took place
during a COVID-19 lockdown, which certainly affected participants’ everyday lives and may have
prevented the realization and formation of certain desired habits to be formed, for example, going
to the gym or indoor areas to do exercise, and (5) similarly, the negative effects of the COVID-19
lockdown may have affected participants’ motivation to work on themselves and improving their
study- or work-related skills.
The authors’ future work includes suggesting and incorporating questions that ask participants
how much their behavior deviated from the intended behavior to gain an insight into the deeper process
of habit-forming and particularly how these directly as well as indirectly affect individual learning
processes. As mentioned by Villalobos-Zuniga and Cherubini (2020), more research is required into
how individual learning processes can be facilitated by design features in apps. Another research
direction would be to look at the individual differences in habit forming with the recent self-report
measure instrument, namely the Creature of Habit Scale, which is a 27-item questionnaire to measure
how individuals differ in habitual responding in everyday life taking into account two aspects of
habits, namely routine behavior and automatic responses (Ersche et al., 2017). This instrument also
combined 20 questions relating to anxiety such as worry, tension, apprehension, and nervousness.
For the authors’ current study, these aspects were beyond the scope of the authors’ research, however,
it would be interesting to examine how individual differences, personality and various factors affect
people’s concerns and anxiety, in the development of habits with a larger sample size and also in a
longitudinal study.
Volume 15 • Issue 2
11
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. Action control. Springer.
Bailey, D. H., Duncan, G. J., Cunha, F., Foorman, B. R., & Yeager, D. S. (2020). Persistence and fadeout of
educational intervention effects: Mechanisms and potential solutions. Psychological Science in the Public Interest,
21(2), 55–97. Advance online publication. doi:10.1177/1529100620915848 PMID:33414687
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-Hal, Inc.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a
fundamental human motivation. Psychological Bulletin, 117(3), 497–529. doi:10.1037/0033-2909.117.3.497
PMID:7777651
Biedermann, D., Schneider, J., & Drachsler, H. (2021). Digital self-control interventions for distracting media:
Systematic review. Computer Assisted Learning, 37(5), 1217–1231. doi:10.1111/jcal.12581
Breitwieser, J., Neubauer, A., Schmiedek, F., & Brod, G. (2022). Self-regulation prompts promote the achievement
of learning goals – but only briefly: Uncovering hidden dynamics in the effects of a psychological intervention.
Learning and Instruction, 80, 101560. Advance online publication. doi:10.1016/j.learninstruc.2021.101560
Brown, J. M., Miller, W. R., & Lawendowski, K. (1999). The Self-Regulation Questionnaire. https://casaa.unm.
edu/inst/SelfRegulation%20Questionnaire%20(SRQ).pdf
Daumiller, M., & Dresel, M. (2019). Supporting self-regulated learning with digital media using motivational
regulation and metacognitive prompts. Experimental Education, 87(1), 161–176. doi:10.1080/00220973.201
8.1448744
Deci, E. L., & Ryan, R. M. (2008). Self-determination theory: A macrotheory of human motivation, development,
and health. Canadian Psychology, 49(3), 182–185. doi:10.1037/a0012801
Edwards, E. A., Lumsden, J., Rivas, C., Steed, L., Edwards, L. A., Thiyagarajan, A., Sohanpal, R., Caton,
H., Griffiths, C. J., Munafò, M. R., Taylor, S., & Walton, R. T. (2016). Gamification for health promotion:
Systematic review of behaviour change techniques. BMJ Open, 6(10), 1–6. doi:10.1136/bmjopen-2016-012447
PMID:27707829
Elaheebocus, S., Weal, M., Morrison, L., & Yardley, L. (2018). Peer-Based Social Media Features in Behavior
Change Interventions: Systematic Review. Medical Internet Research, 20(2), e20. Advance online publication.
doi:10.2196/jmir.8342 PMID:29472174
Ersche, K., Lim, T., Ward, L., Robbins, T., & Stochl, J. (2017). Creature of Habit: A self-report measure of habitual
routines. Personality and Individual Differences, 116, 73–85. doi:10.1016/j.paid.2017.04.024 PMID:28974825
Fiorella, L. (2020). The Science of Habit and Its Implications for Student Learning and Wellbeing. Educational
Psychology Review, 32(3), 603–625. doi:10.1007/s10648-020-09525-1
Galla, B. M., & Duckworth, A. L. (2015). More than resisting temptation: Beneficial habits mediate the
relationship between self-control and positive life outcomes. Journal of Personality and Social Psychology,
109(3), 508–525. doi:10.1037/pspp0000026 PMID:25643222
Hermsen, S., Frost, J., Renes, R., & Kerkhof, P. (2016). Using feedback through digital technology to disrupt
and change habitual behavior: A critical review of current literature. Computers in Human Behavior, 57, 61–74.
doi:10.1016/j.chb.2015.12.023
Lally, P., van Jaarsveld, C., Potts, H., & Wardle, J. (2010). How are habits formed: Modelling habit formation
in the real world. European Journal of Social Psychology, 40(6), 998–1009. doi:10.1002/ejsp.674
Lee, H., Ahn, H., Choi, S., & Choi, W. (2014). The SAMS: Smartphone Addiction Management System and
Verification. Journal of Medical Systems, 38(1), 1. Advance online publication. doi:10.1007/s10916-013-0001-1
PMID:24395031
Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation: A
35-year odyssey. The American Psychologist, 57(9), 705–717. doi:10.1037/0003-066X.57.9.705 PMID:12237980
Volume 15 • Issue 2
12
Daniel Biedermann is a doctoral researcher at the Educational Technologies group at the DIPF, Leibniz Institute
for Research and Information in Education.
Oinas-Kukkonen, H., & Harjumaa, M. (2009). Persuasive Systems Design: Key Issues, Process Model & System
Features. Communications of the Association for Information Systems, 24(1). Advance online publication.
doi:10.17705/1CAIS.02428
Pinder, C., Vermeulen, J., Cowan, B., & Beale, R. (2018). Digital Behaviour Change Interventions to Break and
Form Habits. ACM Transactions on Computer-Human Interaction, 25(3), 1–66. doi:10.1145/3196830
Prochaska, J. O., & Di Clemente, C. C. (1983). Stages and Processes of Self- Change of Smoking - Toward
An Integrative Model of Change. Consulting and Clinical Psychology, 51(3), 390–395. doi:10.1037/0022-
006X.51.3.390 PMID:6863699
Rosenstock, I. M. (1974). The Health Belief Model and Preventive Health Behavior. Health Education
Monographs, 2(4), 354–386. doi:10.1177/109019817400200405
Statista. (2019). Percentage of adults in the United States who use social networks by gender. https://www.
statista.com/statistics/471345/us-adults-who-use-social-networks-gender/
Stojanovic, M., Grund, A., & Fries, S. (2020). App-Based Habit Building Reduces Motivational Impairments
During Studying – An Event Sampling Study. Frontiers in Psychology, 11, 167. Advance online publication.
doi:10.3389/fpsyg.2020.00167 PMID:32116961
Verplanken, B., & Orbell, S. (2003). Reflections on Past Behavior: A Self-Report Index of Habit Strength.
Journal of Applied Social Psychology, 33(6), 1313–1330. doi:10.1111/j.1559-1816.2003.tb01951.x
Villalobos-Zúñiga, G., & Cherubini, M. (2020). Apps That Motivate: A Taxonomy of App Features Based on
Self-Determination Theory. International Journal of Human-Computer Studies, 140, 102449. Advance online
publication. doi:10.1016/j.ijhcs.2020.102449
Available via license: CC BY 3.0
Content may be subject to copyright.
