Research

Methodology
MODULE 2
RESEARCH PROCESS
Objective

 Steps Involved in research process

"The Role of Artificial Intelligence
in Enhancing Healthcare
Diagnostics"
 Problem Statement: Can AI-powered diagnostic tools
improve accuracy and reduce time in detecting diseases
like cancer, compared to conventional methods?

 Literature Review:
Studies highlight AI's success in imaging-based diagnoses,
Gaps identified: Lack of real-world validation, bias in training
datasets, and insufficient focus on physician-AI collaboration.
 Research Questions:
How accurate are AI diagnostic tools compared to traditional methods?
How does AI integration affect the decision-making process in healthcare
settings?
 Variables Identified:
Independent Variables: AI diagnostic tools, traditional diagnostic methods.
Dependent Variables: Diagnostic accuracy, diagnostic time, physician trust.
Control Variables: Patient demographics, type of disease, healthcare setting
Phase II-
 Research Design:
Type: Mixed-methods (quantitative and qualitative).
Approach: Experimental and observational.
 Study Population:
Physicians in oncology and radiology departments, patients undergoing diagnostic procedures.
 Sampling:
Quantitative: Random sample of 1,000 diagnostic cases across hospitals using AI and traditional methods.
Qualitative: Purposive sample of 30 physicians for in-depth interviews.
 Data Collection Tools:
 Quantitative:
Diagnostic reports for accuracy comparison.
Time logs for diagnostic completion.
 Qualitative:
Semi-structured interview guides to assess physician perspectives.
 Ethical Considerations:
Obtain informed consent from patients and physicians.
Ensure AI tools comply with ethical standards and patient data privacy.
 Hypotheses:
AI diagnostic tools are significantly more accurate than traditional methods.
The integration of AI reduces diagnostic time without compromising accuracy.
 Phase III-
 Data Collection:
 Quantitative Phase:

Collect diagnostic results and time metrics for 500 AI-based and 500 traditional diagnostic
cases.
Measure accuracy by comparing results against biopsy-confirmed outcomes.
 Qualitative Phase:

Conduct interviews with physicians about their experience using AI, focusing on perceived
benefits and challenges.
 Data Analysis:
 Quantitative:
Perform statistical tests (e.g., t-tests) to compare diagnostic accuracy and time metrics.
Use regression analysis to examine relationships between patient demographics and AI
performance.
 Qualitative:
Code and analyze interview data to identify themes like trust in AI, usability, and decision-
making challenges.
 Findings:
AI diagnostic tools showed a 10% higher accuracy rate compared to traditional
methods (p < 0.05).
Time for diagnosis reduced by 20% on average when using AI.
Physicians expressed concerns about algorithm transparency but appreciated
AI as a decision-support tool.
 Dissemination:
Publish results in a peer-reviewed medical journal.
Present findings at healthcare conferences and workshops.
Develop policy recommendations for ethical AI implementation in healthcare.
Phases of Research Process

 Phase I:
 Deciding what to research
 Phase II:
 Planning a research study
 Phase IIÏ:
 Conducting a research study
Phase I:
Deciding what to research
 Step I: Formulating a research problem
Phase II:
Planning a research study
 Step II
 Conceptualising a research design
 Step III
 Constructing an instrument for the data collection
 Step Iv
 Selecting a sample
 Step V
 Writing a research proposal
Phase IIÏ:
Conducting a research study
 Step VI
 Collecting data
 Step VII
 Processing and displaying data
 Step VIII
 Writing a research report
 Phase  Phase IIÏ: Conducting a research
I: Deciding what to
study
research
 Step VI
 StepI: Formulating a research  Collecting data
problem
 Step VII
 Phase II: Planning a research  Processing and displaying data
study  Step VIII
 Step II  Writing a research report
 Conceptualising a research design
 Step III
 Constructing an instrument for the
data collection
 Step Iv
 Selecting a sample
 Step V
 Writing a research proposal
Phase I: DECIDING what to do
Step I: Formulating a research problem
 Most important step, because the following steps are influenced by the
research problem
 What do you want to find out about?
 Have you got sufficient funds to do the research?
 Have you got the time available to conduct the study?
 Have you got knowledge of relevant disciplines?
 Do you have sufficient knowledge of skills needed?
Phase II: PLANNING a research study
Step II: Conceptualising a research design
What you find depends on how it was found
 Select an appropriate research design:
 Quantitative
 Qualitative
 Mixed methods
 The design has to be
 Valid
 Workable
 Manageable
 Be aware of its strengths and weaknesses
Phase II: PLANNING - continued

Step III: Constructing an instrument for data collection

How will you collect your data?
 Construct a research instrument or research tool to collect data (interview
schedules, questionnaires, notes on observations, diaries, interview guides,
etc.)
 Or use secondary data (information already collected for other purposes)
 Do a pre-testing of your research tool (pilot study)
Phase II: PLANNING – continued

Step IV: Selecting a sample

Who will take part in your research?
 Select appropriate sample/participants to represent the study population
 Avoid bias
 Random / probability samples
 Non-random / non probability samples
 Be aware of strengths and weaknesses of different sampling methods
Phase II: PLANNING - continued

Step V: Writing a research proposal

Write a detailed plan about your research:
 What are you proposing to do?
 How you plan to proceed?
 Why you select the proposed strategy?
Phase III: CONDUCTING a research study

Step VI: Collecting data

Doing the data gathering using one or more data collection method,
such as:
 conducting interviews
 mailing out questionnaires
 conducting nominal/focus groups discussions
 making an observation

Be aware of ethical issues!

Phase III: CONDUCTING - continued

Step VII: Processing and displaying data

What did you find how?
 Analysing the data depends on the type of information and how to
communicate the findings
 Distinguish between
 Descriptive
 Quantitative (statistical procedures)
 Qualitative
Phase III: CONDUCTING - continued
Step VIII: Writing a research report
What have you done? What conclusions have you drawn from the
findings?
 Different format for quantitative and qualitative research
 Structure using main themes of study
 Use academic conventions
Thank You