IEEE Technology and Society Magazine - September 2023 - 78

Feature
Some researchers differentiated GT from content
analysis [22] as two distinctive research methods.
Rather than contrasting them, we adapt the GT in
this study as a research design framework, mainly
to inform and guide the content analysis. In general,
the content analysis consists of four stages:
data collection, coding, analysis, and interpretation
of coded content [16]. In the data collection
stage, researchers select data sources and identify
sampling criteria. In the coding phase, collected
textual data are coded into different categories at
various levels such as words, phrases, sentences,
paragraphs, or themes. The analysis stage involves
various ways and formats to present codes and relationships
among different code segments. In the GT
framework, the coding phase and analysis phase
are intertwined to a certain extent. The authors of
this study took on relational analysis that involves
exploring the relationships between concepts. An
evaluation of the co-occurrence of explicit concepts
in the text was performed to gain insights and a big
picture of the phenomenon. The interpretation stage
involves the interpretation of results and drawing
logical and reasonable conclusions.
Data collection
One unique attribute of GT lies in its iterative and
overlapping processes between data collection and
data analysis [23]. To incorporate such design into
the content analysis process, the authors started with
initial data collection based on purposive sampling
strategy. The initial search of keywords " Artificial
Intelligence, " " Healthcare, " and " Ethics " from the
online databases including PubMed, ProQuest, IEEE
Xplore, and JSTOR returned 31,193 results. Resource
types included academic articles, newspaper articles,
books, reports, and even video resources. The
authors also attempted several different combinations
and variations of keyword synonyms and
Table 1. Synonyms of search strings.
retrieved another around 40,000 results. The search
was conducted on titles, abstracts, and keywords.
The search terms were organized into three distinct
groups: " artificial intelligence, " " healthcare, " and
" ethical principles. " The synonyms for each of the
three terms are listed in Table 1. Subsequently, we
systematically selected one search term from each
group, creating sets of three search terms. Each set
of search terms always contained one word from
each group. For instance, " AI, " " Health, " and " Ethical
values " could form a set of search strings used
to produce database search results. By exhaustively
combining search terms from all three categories, we
generated various sets to facilitate comprehensive
exploration. The search was conducted using the following
expression: [ " artificial intelligence " OR " AI "
OR " computational intelligence " OR " intelligent system "
OR " smart technology " ] AND [ " healthcare " OR
" health care " OR " healthcare provision " OR " healthcare
service " OR " medical care " OR " medical maintenance "
OR " medical service " OR " medical support "
OR " medical treatment " ] AND [ " ethics " OR " code of
ethics " OR " ethical guidelines " OR " ethical standards "
or " ethical principles " OR " ethical values " ].
After the initial screening of data source types,
the number of data sources decreased to 17,945.
Since one of the purposes of the study is to analyze
the trend of development of AI ethical principles,
all data sources must be recent (within the past
10 years from 2022) and notably address specific
topics regarding ethical regulations of AI in healthcare.
The specified time range further reduced the
number of retrieved documents to 16,481. The goal
at this stage was to select data sources containing
discussions about all three topics intended to be
studied: AI, healthcare, and ethics. In addition,
credibility of the data sources was also evaluated
and screened. Only peer-reviewed academic articles
and reports published by governments and
78
IEEE Technology and Society Magazine
IEEE Technology and Society Magazine - September 2023

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - September 2023
