Unit2 AB

UNIT II | Research Design and Methodology

1. Research Design

• A research design is a structured plan or framework guiding the conduct of a research study.

• It details methods for collecting, measuring, and analyzing data to answer a specific research question effectively.

• Purpose: Enables systematic investigation of research objectives while minimizing bias and allowing for valid and reliable conclusions.

Characteristics of a Good Research Design:

• Objectivity: Free from personal bias.

• Reliability: Produces consistent results under similar conditions.

• Validity: Measures what it is intended to measure.

• Flexibility: Can adapt to minor procedural changes.

• Economical: Uses time and resources efficiently.

• Ethically Sound: Ensures participant safety and confidentiality.

Functions of Research Design:

• Provides a plan of action for data collection and analysis.

• Ensures valid and reliable results.

• Reduces biases and errors.

• Helps in budgeting and resource allocation.

• Enables replication by other researchers.

Components of Research Design Outline:

• Type of data to be collected (qualitative or quantitative).

• Methods and instruments for data collection.

• Timeline and setting of the study.

• Criteria for selecting subjects or samples.

• Procedures for data analysis.

2. Types of Research Designs

Exploratory Research Design

• Used when seeking insight or familiarity with a phenomenon or problem when little information exists.

• Purpose:

  • Formulate a research problem or hypothesis.

  • Explore new ideas and concepts.

• Methods include:

  • Literature surveys

  • Case studies

  • Expert interviews

  • Focus group discussions.

• Example: Studying the attitudes of patients toward a newly introduced telemedicine service before a large-scale trial.

Descriptive Research Design

• Aims to describe characteristics of a population or phenomenon accurately.

• Answers 'what', 'where', 'when', and 'how' but not 'why'.

• Purpose:

  • Provide a detailed and factual picture of a situation.

  • Measure variables and describe relationships between them.

• Types of Descriptive Designs:

  • Cross-sectional study: Data collected at one point in time.

  • Longitudinal study: Data collected over a period of time.

  • Case study: In-depth analysis of a single case or group.

• Example: A survey assessing the prevalence of hypertension among adults in a particular city.

Experimental Research Design

• Used to test cause-and-effect relationships between variables under controlled conditions.

• Purpose:

  • Determine the impact of an independent variable on a dependent variable.

  • Test hypotheses through manipulation and control.

Types of Experimental Designs:

• Randomized Controlled Trial (RCT):

  • Scientific study where participants are randomly allocated to different groups (e.g., treatment and control) to minimize bias and confounding factors.

  • Designs included:

  • Parallel group design

  • Crossover design

  • Cluster randomized trial

  • Adaptive RCT.

• Factorial Design:

  • Examines multiple independent variables and their interaction effects.

  • Example: Testing two dosages of a drug and two types of diets to see combined effects.

• Latin Square Design (LSD):

  • Experimental design used to control two sources of variability (nuisance variables) in addition to the treatment effect.

• Post-test only control group design:

  • Participants are randomly assigned to treatments or control groups, and post-test results are compared.

Other Experimental Designs:

• Pre-Experimental Design:

  • Used for preliminary testing with minimal control over variables.

  • Examples:

  • One-group pre-test post-test design: Same group measured before and after treatment.

  • One-shot case study design: A single group exposed to treatment and outcome measured afterward.

• True Experimental Design:

  • Uses randomization, control group, and manipulation of variables.

• Quasi-Experimental Design:

  • Used when randomization is not possible.

  • Examples:

  • Non-equivalent control group design.

  • Time-series design: Repeated observations before and after the intervention.

  • Example: Measuring hospital infection rates before and after implementing a new sanitation policy.

3. Sampling in Research

• Definition: Sampling is the process of selecting a subset (sample) from a larger group (population) to collect data.

• Allows for valid inferences about the entire population with limited resources.

Importance of Sampling:

• Feasibility & Cost-effectiveness: Reduces cost, time, and effort.

• Efficiency: Accelerates research process.

• Accuracy & Precision: Well-conducted samples can yield accurate results.

• Enables Intensive Study: Focus on high-quality data collection.

• Statistical Inference Basis: Estimates population parameters and assesses reliability of results.

• Ethical Considerations: Minimizes participant burden and exposure risks.

Factors Influencing Sample Size Determination:

• Effect Size: Magnitude of the difference or relationship to detect.

• Study Objectives: Purpose of study affects sample size.

• Population Size: Larger populations require different sample considerations.

• Variability of Population: More diverse populations require larger samples.

• Significance Level (α): Probability of Type I error; smaller α requires larger samples.

• Confidence Level: Higher confidence levels require larger samples.

• Statistical Power: Higher power requires a larger sample size.

• Study Design: Different designs yield different sample requirements.

• Drop-out Rate: Anticipated participant loss may inflate initial sample size.

4. Sampling Designs: Probability vs. Non-Probability

Probability Sampling:

• Every individual in the population has a known, non-zero, equal chance of selection.

• Key Features:

  • Random selection process

  • Known inclusion probability

  • Allows calculation of sampling error

  • Results are generalizable to the population.

Types of Probability Sampling:

• Simple Random Sampling: Equal chance of selection; done with random number tables/software.

• Systematic Sampling: Select every kth element from an ordered list.

• Stratified Sampling: Population divided into subgroups, samples taken from each.

• Cluster Sampling: Population divided into clusters; entire clusters are sampled.

• Multistage Sampling: Combination of several probability methods in stages.

Non-Probability Sampling:

• Selection based on researcher’s judgment or convenience; not all individuals have a known chance of selection.

• Key Features:

  • No randomization

  • Subject to sampling bias

  • Results are less generalizable.

Types of Non-Probability Sampling:

• Convenience Sampling: Subjects selected based on availability.

• Purposive Sampling: Selected based on specific characteristics.

• Quota Sampling: Specific subgroups represented, but non-random within each quota.

• Snowball Sampling: Existing subjects recruit future participants.

• Consecutive Sampling: All subjects meeting criteria are selected until desired sample size is reached.

Comparison Table: