High analytical and statistical technologies are
contrasted, accordingly, with low analytical and
statistical technologies (with classical ones
situated between them). “Low analytical and
statistical technologies” are those technologies
that do not correspond to the modern level of
science and practice. Usually, they are
simultaneously outdated and not entirely
adequate to the tasks being solved.
Classical analytical and statistical technologies
are understood as long-standing technologies that
have retained their importance for modern
analytical and statistical practice. Such
technologies based on the method of least
squares (including methods of point estimation
of parameters of the predictive function, non-
parametric methods of confidence estimation of
parameters and the predictive function as a
whole, tests of various hypotheses about them),
Kolmogorov, Smirnov, omega-square type
statistics, non-parametric Spearman and Kendall
correlation coefficients (to attribute them only to
ranking analysis methods means to condescend
to “low analytical and statistical technologies”)
and many other statistical procedures.
There is another approach (Köbis, Starke, &
Rahwan, 2022; Kovtun, 2011; Odilla, 2023) to
the classification of analytical and statistical
technologies, according to which they can be
divided into traditional (classical) and
automated. The classical one involves the
activities of 1-2 experts with a minor application
of computer (information) technologies, during
which analysis of interrelations,
interdependencies of various indicators is carried
out to identify deviations from the norm (this
includes stereotype methods, adjusted indicators,
associated comparisons, and so on).
In turn, automated group assumes, firstly, the
involvement of intelligent systems in the data
processing, which are trained to perform
analytical and statistical operations with datasets
and capable of replacing a person in most
operations performed. These technologies
include data mining, anomaly detection in them,
or new knowledge discovery (novelty detection,
knowledge discovery). The use of these
algorithms allows the automation of the work of
several specialists and simplifies the process of
preparing reports about exceptional (new,
anomalous) situations. Based on this
classification, automated technologies should be
considered as innovative analytical and statistical
technologies. We can add that the use of the
indicated technologies in modern conditions of
societal development is impossible without the
use also of such technologies as machine
learning, deep learning, neural networks, natural
language processing (NLP), network analysis,
real-time data analysis, which allow obtaining
fast and accurate results of analysis of large
volumes of data.
As noted by (Köbis, Starke, & Rahwan, 2021),
the active implementation of artificial
intelligence and machine learning technologies
brings new hope for more effective corruption
counteraction. Artificial intelligence differs
significantly from static information and
communication technologies. “Classic”
technologies allow to digitize the procurement
procedures, the provision of government services
online, and the publication of open government
data. However, traditional technologies cannot
operate autonomously, while artificial
intelligence, on the contrary, was specifically
designed for this. Thanks to its learning
capabilities, artificial intelligence can
autonomously perform a wide range of tasks
previously reserved for humans (Rahwan et al.,
2019). In the context of today's digital
transformation of states, artificial intelligence
can take on anti-corruption tasks, such as
predicting, detecting, and exposing corruption
cases (Lima & Delen, 2020; López-Iturriaga &
Sanz, 2017).
Thus, in practical use, applied statistical methods
and analytics involve not just separate data
description methods, estimation, hypothesis
testing, but complete, integrated procedures – so-
called “analytical and statistical technologies”.
The concept of “analytical and statistical
technology” in our understanding is analogous to
the concept of “technological process” in the
theory and practice of production organization
(The Ukranian Week, 2022).
Naturally, some statistical technologies better
meet the needs of the researcher (user,
statistician), others worse; some are modern,
others outdated; the properties of some are
studied, others not. It should be noted that
competent and effective application of analytical
and statistical methods is by no means just a
check of one separately taken statistical
hypothesis or an assessment of characteristics or
parameters of one given distribution from a fixed
family. Such operations are only a separate brick
that makes up the analytical and statistical
technology. The procedure of analytical and
statistical data analysis is an information
technology process, i.e., a certain information
technology in which statistical information