With the right systems in place, collection
of data is more likely to efficiently lead to
quality information.
On the other hand, correct research
design, instruments and methodologies
form the other integral part of research;
with the right skills, which is readily
available in the market now, this is no
longer a big problem. One remaining part
is the downstream; the ability to extract
insights and new knowledge from research
to improve prior knowledge on different
states of nature.
Some years back, in an industry discussion
“the gathering” by Financial Services
Deepening Trust (FSDT) Tanzania, one
of the participants observed that in many
organizations, as much as research surveys
are commissioned and data collected, the
use of the information is very minimal.
Decisions mainly based on senior
managers’ gut feelings and experiences.
It was observed that one of the
contributing causes for this is lack of
capacity to comprehend research results.
In most cases the results are interpreted
using the understanding of the market
research professional, full of expert’s
terminologies which sometimes does not
necessarily conform to the reality in the
organization and its strategy.
One area of improvement is the way the
information is presented. According to the
work of psychologists Daniel Kahneman
and Amos Tversky (Daniel Kahneman,
2012) on fast and slow thinking, our brains
have a limited capacity in holding and
processing multidimensional information.
In addition, while making decisions, we
are also affected by heuristics which result
in cognitive biases; this is the tendency to
draw an incorrect conclusion in a certain
circumstance based on cognitive factors
which affects our ability to objectively
make decisions, in turn, it affects the value
of market research findings.
Heuristics are simple, efficient rules which
people often use to form judgments and
make decisions. They are mental shortcuts
that usually involve focusing on one aspect
of a complex problem and ignoring others.
Some of the common heuristics are
anchoring (Interpreting information based
on known facts), framing (the style with
which the information is presented to the
decisions maker), availability (vivid, easily
38 MAL34/20 ISSUE
Data blending, integration and visual-
ization aids in better conceptualizing, it
helps our mind to decipher information
faster and see the complete picture. In
cases of multiple people being involved
in the process at the same time, it helps
in ‘seeing things in the same way’, which
can reduce framing bias.
imagined, but uncommon events are
highly weighted in brains), Confirmation
(initial decisions being self-fulfilling
prophesies), commitment escalation
(difficulty in allocating resources where
reward is not guaranteed) and hindsight
bias (Difficulty in remembering when we
didn’t know what we know, which hinders
our ability to learn from past mistakes).
Data blending, integration
and visualization
One way of overcoming such biases
and making better decisions, which
may result in increasing the expected
monetary value of research findings, is
considering many different variables at
the same time, otherwise called features.
It can be achieved through blending and
integrating different types of data from
different sources and being able to view
them concurrently.
Data
blending,
integration
and
visualization
aids
in
better
conceptualizing, it helps our mind to
decipher information faster and see the
complete picture. In cases of multiple
people being involved in the process at
the same time, it helps in ‘seeing things in
the same way’, which can reduce framing
bias.
Integration completes the picture. In this
case, different data sets from different
sources such as organizations’ internal
information; finance, marketing or even
budgeting data, are integrated. This helps
in reducing anchoring bias. For example,
integrating survey data into the company
information would lead to better business
intelligence about products, customers,
the market, competition etc.
A good dashboard system can be used
together with this information in a proper
format for mining insights from the data.
Conclusion
Theoretically, knowledge-based decision
making underpins every successful
organization. But, as Plato pointed out,
“Human behavior flows from three main
sources: desire, emotion, and knowledge”
(Borg, 2015). When we make decisions
based on right knowledge, we are more
likely to reduce our biases in desire and
emotions.
References
Borg, M. A. (2015). Changing behaviour
to achieve more effective infection
control compliance. J Patient Saf Infect
Control Elsevier, 3, 27–28.
Kahneman, D, & Tversky, A. (1979).
Prospect theory: An analysis of decision
under risk. Econometrica, 47, 263–291.
Kahneman, Daniel. (2012). Thinking,
fast and slow. Penguin Books.
Charles Makau is a data science
expert,
particularly
in
big
data analytics, integration and
visualization. He is a director at
African Stats, an integrated custom
research, data driven strategies and
training outfit. You can engage him
via email at: Charles.Makau@
africanstat.com.