Research Design, MLS 394

• Research Design I, MLS 394, Lecture date: 28/03/2025, Latest update: 28/03/2025, Dr. Abdulelah Alshawli, [email protected].

Learning Outcomes

• Understand the concept of research design
• Identify and differentiate between types of research designs
• Recognize the importance of research design in successful research
• Evaluate real-life examples of successful and failed research designs

What is Research Design?

• Definition: A plan outlining how to conduct research to ensure proper data collection and analysis
• Importance:
  • Helps answer research questions accurately
  • Minimizes bias and errors
  • Facilitates reproducibility
• Good Design Example: A study testing a new glucose meter on a group of patients, using both the new glucose meter and a standard meter allows for a direct comparison
• Bad Design Example: A study testing the new glucose meter on a group of patients with no control group, making it not possible to determine whether the measurements are correct or not

Components of Research Design

• A clear, specific and measurable research question. Example: "How accurate is a new glucose meter compared to a standard one for diabetic patients?"
• Hypothesis Formulation: A testable hypothesis based on the research question. Example: "The new glucose meter gives readings within 10% of lab results more often than a standard meter."
• Appropriate Methodology: The research method aligns with the goals. Example: A trial where patients use both meters, and results are compared
• Sampling Plan: A clear strategy for selecting participants. Example: 100 diabetic patients from KAUH diabetic clinic, diverse in age and gender
• Data Collection Techniques: Use reliable and valid tools for collecting data. Example: Measure glucose using the new meter and a standard lab test
• Control Variables: Identifying and managing factors that could affect the results. Example: Test at the same time daily, 16 hours after fasting
• Ensuring ethical standards are met in research. Example: Get patient consent, keep data private
• Outline methods for analyzing collected data. Example: Compare meter readings with a statistical test (e.g., paired t-tests)
• Replication and Reliability: Designing studies that can be replicated. Example: Providing detailed methodology so others can replicate the study
• Limitations and Delimitations: Acknowledging the study's limitations and scope. Example: The study only includes KAUH diabetic clinic patients, may not apply everywhere
• Clear Reporting: Transparent reporting of results and methodology. Example: Publish findings with a detailed methods section, results, and discussions in a peer-reviewed medical journal

Types of Research Design Include

• Quantitative Research Designs (Descriptive, Correlational, Experimental, Quasi-Experimental, Longitudinal, Cross-Sectional, Retrospective, Prospective)

• Qualitative & Mixed-Methods (Case Study, Mixed-Methods)

• Other Designs (Pilot Study, Non-Experimental)

• Descriptive Research Design observes and describes a population, phenomenon, or behavior without manipulating variables or testing cause-and-effect. Example: Measures the percentage of diabetic patients in KAUH diabetic clinic who use the new glucose meter

• Correlational Research Design examines the relationship between two or more variables. Example: Checks if frequent use of the new meter relates to better glucose control

• Experimental Research Design tests the effect of changing one or more independent variables (IV) to observe effects on a dependent variable (DV). Example: Tests if using the new glucose meter (IV) provides accurate readings (DV) closer to the standard lab test, while keeping measurement timing, patient conditions, and procedures constant (constant variable)

• Quasi-experimental Research Design: Similar to the experimental design but lacks random assignment. Example: Compares glucose readings from patients using either meter, without random assignment (male patients only)

• Longitudinal Research Design involves repeated observations of the same variables over time. Example: Tracks glucose levels in patients using the new meter over two years

• Cross-sectional Research Design observes a population at a single point in time. Example: Tests the new meter's accuracy at one health screening event (students' activity)

• Retrospective Research Design has the purpose of analyzing existing data collected in the past to identify trends or outcomes. Example: Looks at past clinic data to compare new meter readings to old ones

• Prospective Research Design has the purpose of following participants, forward in time to observe outcomes or changes. Example: Follows patients using the new meter for six months to see improvements

• Case Study is an in-depth analysis of a single case (individual, group, or event) to explore a phenomenon in real-world context. Example: Tracking a single diabetic patient's experience using the new glucose meter over six months, analyzing accuracy, usability, and impact on blood sugar management compared to standard lab tests

• Mixed-Methods combines qualitative and quantitative research methods to provide a more comprehensive understanding of a research question. Example: Combines meter accuracy data with patient feedback interviews

• Pilot Study: A small-scale preliminary study conducted to evaluate feasibility, time, cost, and adverse events involved in a research project. Example: Tests the new meter with 20 patients to check for usability and reliability among a small group of diabetic patients before a larger clinical trial

• Non-Experimental research design aims to observe and describe variables without manipulation, a method often used in descriptive, correlational, or observational studies. Example: Observes how patients feel about the new meter's ease of use

Factors to Consider in Research Design

• Clear Research Question and Objectives: Set specific, measurable, achievable, relevant, and time-bound (SMART) objectives (e.g., How accurate is the new glucose meter?)
• Type of Data Needed: Identify whether you need quantitative data (numerical readings) or qualitative data (patient experiences). Choose appropriate data collection methods (surveys, measurements, etc.)
• Resources Available: Assess funding, personnel, equipment, and time constraints. Ensure access to participants and necessary facilities
• Ethical Considerations: Obtain informed consent and ensure participant confidentiality. Minimize risks and review the study with an ethics committee

Common Pitfalls in Research Design

• Lack of Clarity can present problems with accurately interpreting, understanding, and applying research data. Example: Asking, "Is the new meter is good?" without defining what "good" means
• Choosing the Wrong Design can lead to inaccurate conclusions about the underlying research question. Example: Using patient opinions to test meter accuracy instead of lab data
• Insufficient Sample Size can lead to unreliable results. Example: Testing only 10 patients, missing the full picture of the meter's accuracy

Understand research design concepts and differentiate between types of designs. Recognize its importance in ensuring accurate data collection, minimizing bias, and facilitating reproducibility. Evaluate real-life examples of successful and failed research designs to enhance learning.