Content uploaded by Kyriakos Tsoukalas
Author content
All content in this area was uploaded by Kyriakos Tsoukalas on May 17, 2021
Content may be subject to copyright.
Self-Guiding Behavior Change using a Mobile App
Kyriakos D. Tsoukalas
Human Centered Design
Virginia Tech
ktsoukalas@vt.edu
ABSTRACT
Based on the expanding usage of smartphones as
the platform for computer applications, this paper
addresses the information asymmetry between
personal and factual observations that inform a
decision-making process. The purpose of this pa-
per is to propose a model to help self-guided be-
havioral change based on collection of personal
information. A mobile application exemplifies
self-guiding behavioral change through the col-
lection of daily quantitative data and the retrieval
of descriptive statistics during decision-making.
INTRODUCTION
The widespread use of the smartphone introduced
a behavioral change in daily decision making.
Many people use smartphones to store, retrieve,
and search for information. Smartphones com-
municate with a variety of services over the inter-
net, and the applications used create a plethora of
metadata about how people use them (Azar 2003;
van Velsen, 2013). The internet is a medium for
widespread data distribution and collection,
which has been proposed as a medium for behav-
ioral interventions (Ritterband, 2009).
People’s perspectives about the use of mobile ap-
plications to achieve health related behavior
changes is understudied according to Dennison et
al. (Dennison, 2013). Taking advantage of smart-
phones’ growing computational power and mobil-
ity in order to self-guide behavioral changes, re-
quires that the technology allows storage and re-
trieval of data in ways we deem helpful to deci-
sion making.
Kahneman and Tversky developed their prospect
theory to analyze decision making by considering
risks (Kahneman, 1979). The authors posit that
expectations drive people’s perspective in charac-
terizing changes either as gain or as loss. People
have different attitudes, each of which alters
judgement differently. The difference between
what we think is going on and what is actually
going on is information asymmetry. The concept
of information asymmetry is related to transac-
tions between agents that possess different levels
of information.
In the context of decision making, information
asymmetry manifests as a lack of knowledge that
alters the perception of the situation. People often
address a lack of information with information
from past experience if available. In other words,
people link previously experienced prospects to
similar prospects that are under consideration.
Hadar and Fox investigated how decisions differ
based on whether described conditions have been
experienced by the decision maker before or not
(Hadar, 2009). The authors found that decision
making outcomes based on descriptions of pro-
spects that have been previously experienced will
differ from decision making outcomes based on
prospects that have not been previously experi-
enced. Such divergence between decision making
with and without prior experience is attributed to
the information surplus on prospects due to hav-
ing prior experience.
This paper is concerned with personal develop-
ment from a self-guiding point of view. Self-
guiding behavior depends on the readiness of a
person to change habits, and self-perspectives
about which changes are required for the desired
outcome. Boyatzis and Akrivou discuss how per-
spectives of the ideal self, transform into visions
of personal development (Boyatzis, 2006). How-
ever, having personal visions and even increased
awareness of what are causes of personal prob-
lems are not enough to bring about change (Pro-
chaska, 1992). Self-guiding behavior change
needs to assess the prospects of the attempt to
change in order to prepare for different stages of
changing behavior. Blanchard, Zigarmi and Nel-
son discuss the evolution of leadership theories
and in particular the situational approach to lead-
ership, which is relevant to self-guided personal
development (Blanchard, 1993). The authors re-
flect on the changes between the first and second
models of situational leadership. Specifically, the
authors point out that the first situational leader-
ship model categorized people in levels of “em-
ployee development” and the first level included
people falsely assumed to be “unwilling and una-
ble” to perform their work tasks. While in the re-
vised (second) model of situational leadership,
the first level of employee development includes
people who are assumed to be “low in compe-
tence” and “have high commitment” (Blanchard,
1993, p. 27).
The current paper equates employee development
with personal development. Behavior change is a
process of both contemplation and action. Con-
templation is necessary for the alteration of ex-
pectations and self-representations, while com-
mitment is necessary to maintain action towards
change.
The second personal development level is being
competent but having low commitment. The cur-
rent research suggests computing, here the use of
an application on a smartphone (mobile) plat-
form, enables people to collect self-data and
compute descriptive statistics. However, such
competence development is not possible without
being committed to the development of such
skills. This paper’s proposed model labels the be-
havioral development from the first to the second
personal development level as computational
competence development.
The third personal development level is being
highly competent and having variable commit-
ment. People’s experience alters their expecta-
tions and perspectives when considering previ-
ously experienced prospects. There is an infor-
mation asymmetry between decision making with
previously experienced and with novel prospects.
This paper’s proposed model labels the behavior-
al development from the second to the third per-
sonal development level as experience develop-
ment.
The fourth personal development level is being
highly competent and having high commitment.
Maintaining a high level of behavioral commit-
ment after having developed high computational
competence depends on the expectations for, and
perspectives on the goals at hand, which are mod-
erated by prior experience. Simply, self-
motivation empowers commitment. Motivation is
an affective state driven by people’s expectations
and their different self-perspectives. Motivation
manifests as behavioral repetition driven by
commitment in a specific behavior. The proposed
model labels the behavioral development from
the third to the fourth personal development level
as behavioral commitment development.
PROPOSED MODEL
Figure 1 depicts a model for self-guiding behav-
ior change based on the development of computa-
tional competence. The proposed model is based
on two hypotheses. The first is that prospect
awareness and computational competence are
positively correlated, and the second is that prior
experience and information asymmetry between
goal expectations and prospect awareness are
negatively correlated.
People set goals based on their expectations and
self-perspectives. People’s experience alters ex-
pectations and self-perspectives; thus, the devel-
opment of experience reduces the information
asymmetry between the consideration of previ-
ously experienced prospects and novel prospects.
The ability to collect self-behavioral data to in-
form expectations and self-perspectives addresses
information asymmetries and discrepancies that
regulate prospect awareness. Prospects are judged
in order to make choices.
Decision-making in real-world situations is usual-
ly concerned with multiple objectives that may be
conflicting with each other to various degrees.
Moreover, rationality in decision-making is mod-
ulating based on situational and personal factors
(Kahneman, 2003; Campitelli, 2010). The situa-
tional leadership theory was briefly discussed in
combination with the principle of information
asymmetry due to prior experience.
This paper proposes a computer-aided, via a mo-
bile application, self-guiding behavioral change
model that follows the prescriptive analysis ap-
proach. Proposed design needs for a mobile ap-
plication are described and to the most part im-
plemented in a working prototype with the aim of
conducting future empirical research with users
of the application to inform design needs and col-
lect feedback as well as feature requests.
PROPOSED MODEL
Fig 1. Self-Guiding Behavior Change Model.
Design Needs
• A taxonomy of data-labels that is editable
by the user.
• The user can use a date picker to store or
change a daily data value per label.
• The user can use a date picker to visit a
view of descriptive statistics per month.
The user is able to navigate in time and
among the types of statistics.
• A list of different profiles allows for mul-
tiple user-profiles to be used for data col-
lection.
• Encrypted data repository.
• No internet access needed.
• Import from and export to CSV data files.
• A list of different views allows the differ-
entiation of the graphical user interface to
enable different types of data collection
(i.e., health, financial, household, etc.)
Design Limitations
• Only one value per day per data-label can
be stored.
• Descriptive statistics are computed and
presented per calendar month.
Design Implementation
Fig 2. Example Screenshots of the User Interface.
Automated Descriptive Statistics
• Daily records: Each daily value per
month.
• Difference from goal: The difference (%)
of each daily value from a predefined goal
value.
• Tendency: Smoothed out daily values per
month.
• Difference from average: The difference
of each daily value of the month from the
month’s average.
• Progressive average difference: The dif-
ference of each daily value from the aver-
age of daily values from the beginning of
each month up to each daily value of the
month.
DISCUSSION
Research Needs
After the development of the proposed applica-
tion reaches the beta stage, then the collection of
empirical data is needed to investigate how often
users store and browse data, as well as their tax-
onomies of data-labels. Such empirical evidence
will inform a human-centered approach to the de-
sign of the user interface so as to offer multiple
views and alternative functionality tailored to dif-
ferent types of data collection.
Features Under Development
The current features under development include:
1) a machine learning algorithm based on naïve
Bayes classifiers for active learning of data-label
statistical salience, and
2) an alternative function that allows to record
multiple datapoints per day and a statistical view
that presents datapoints on a 24-hour chart.
Contribution
This paper contributes a self-guiding, computer-
aided behavior change model to the literature of
personal development by building on the con-
struct of employee development in the theory of
situational leadership, and on the construct of in-
formation asymmetry between decision-making
of previously experienced and novel prospects.
To enable future research, a prototype mobile ap-
plication was developed to allow for downloading
and testing:
https://play.google.com/apps/testing/com.labsloft.
healthtrack
An implication for the situational leadership theo-
ry is that employee development is essentially
personal development. However, leaders ought to
provide employees with necessary tools for the
work tasks (developing competence), and place
employees in roles that result in self-motivated
work (developing commitment).
Conclusion
In conclusion, people can increase their cognitive
abilities by using computers and software for data
storage and analysis. However, free or low-cost
tools tend to serve their developers’ intention to
collect data from people’s use of their applica-
tions. Self-guiding behavioral change is possible
when people are self-motivated to pursue change.
However, self-collection of behavioral data can
reveal discrepancies between self-representations
of and actual behavior. The accuracy of factual
information during decision making is critical to
people’s awareness of possible prospects, because
misrepresentations of self-data obscure the con-
sideration of possible solutions as well as the con-
templation on the causes of problem.
REFERENCES
1. Azar, K. et al. (2013). Mobile Applications
for Weight Management Theory-Based Con-
tent Analysis. American journal of preventive
medicine. 45. 583-9.
2. Boyatzis, R. E., & Akrivou, K. (2006). The
ideal self as the driver of intentional
change. Journal of Management Develop-
ment, 25(7), 624–642.
3. Campitelli, G. & Gobet, F. (2010). Herbert
Simon’s Decision-Making Approach: Investi-
gation of Cognitive Processes in Experts. Re-
view of General Psychology. 14. 354-364.
10.1037/a0021256.
4. Dennison, L. et al. (2013). Opportunities and
Challenges for Smartphone Applications in
Supporting Health Behavior Change: Qualita-
tive Study. J Med Internet Res 2013;
15(4):e86.
5. Higgins E. T. (1987). Self-discrepancy: a the-
ory relating self and affect. Psychological re-
view, 94(3), 319–340.
6. Kahneman D. (2003). A perspective on judg-
ment and choice: mapping bounded rationali-
ty. The American psychologist, 58(9), 697–
720.
7. Kahneman, D., & Tversky, A. (1979). Pro-
spect Theory: An Analysis of Decision under
Risk. Econometrica, 47(2), 263-291.
8. Liat, H., & Craig, R. F. (2009). Information
asymmetry in decision from description ver-
sus decision from experience. Judgment and
Decision Making, 4(4), 317–325.
9. Prochaska, J., Diclemente, C., & Norcross, J.
(1992). In search of how people change. Ap-
plications to addictive behaviors. The Ameri-
can psychologist, 47 9, 1102-14.
10. Ritterband, L. M. et al. (2009). A behavior
change model for internet interven-
tions. Annals of Behavioral Medicine, 38(1),
18–27.
11. van Velsen, L., Beaujean, D.J. & van Gemert-
Pijnen, J.E. (2013). Why mobile health app
overload drives us crazy, and how to restore
the sanity. BMC Med Inform Decis Mak 13,
23.