Problemas y Diseños de Investigación Cuantitativa - MICS UGR

Introduction to Research Methodology in Health Sciences

Overview of Quantitative Research Designs

• María Eugenia Vila introduces the first class on research methodology, focusing on quantitative research designs prevalent in health sciences.

• The importance of decisions made by researchers regarding the development and approach to addressing research problems is emphasized.

Project Submission Requirements

• Institutions require a project proposal before conducting research, which includes information about the problem and methods to be applied.

• Students must develop and present a thesis plan prior to their final work, highlighting the significance of structured planning in research.

Structure of Research Articles

• A quick overview of how to navigate through a full-text article is provided, including publication data such as authorship, title, journal abbreviation, edition, volume, and pages.

• Keywords are crucial for database searches; they help readers locate articles effectively similar to Google searches.

Components of Full-text Articles

• The introduction section contains theoretical frameworks, justifications for the study, problem statements, and objectives.

• The methods section details study population information and intervention procedures; it may also include statistical analysis plans.

Results and Discussion Sections

• Results are presented narratively with statistical data and graphics; this section describes findings from the study.

• The discussion analyzes results against other studies; conclusions summarize key findings related to study objectives while suggesting future research directions.

Understanding Research Problems

Types of Research Problems

• In previous coursework (introduction to methodology), students identified specific problems for investigation. This course will delve deeper into analyzing these problems.

Categories of Problems

• Four types of problems are distinguished: descriptive, associative/correlational, comparative, and explanatory. Each type dictates appropriate methodological approaches.

Descriptive Problems

• Descriptive problems aim to understand variable distributions within populations without exploring relationships or causations.

• Example: Investigating physical activity patterns in Chilean populations focuses solely on description rather than correlation with health outcomes.

• Another example involves assessing clinical characteristics in children with severe bronchiolitis without linking them to treatment outcomes.

Understanding Variable Relationships in Research

The Importance of Data Collection

• To analyze relationships between variables, data must be collected at multiple points in time. This can be likened to taking a series of photos or videos to capture changes over time.

Hypothesis Formation from Descriptive Studies

• Descriptive studies can lead to hypotheses suggesting that two or more variables may be related. Changes in the independent variable could imply changes in the dependent variable, which is crucial for understanding associations.

Predictive Power of Identified Relationships

• If a correlation is found (e.g., higher absenteeism among those living peripherally), interventions can be implemented to improve accessibility and address issues effectively. This predictive capability allows for proactive measures based on research findings.

Intervention Effects Evaluation

• Once relationships are established, it becomes essential to evaluate the effects of specific interventions (e.g., vibratory therapy on patients with multiple sclerosis). This involves assessing outcomes before and after intervention application.

Comparative Analysis of Diagnostic Methods

• Comparing different diagnostic methods' sensitivity, specificity, and predictive power is vital for determining their effectiveness against established standards (gold standard). Such comparisons help refine diagnostic practices in healthcare settings.

Designing Effective Research Studies

Initial Steps in Research Design

• When initiating research (e.g., establishing kinesiology services), understanding the population composition is critical. A descriptive study may reveal significant trends, such as high demand for respiratory care among children during winter months.

Correlation Between Socioeconomic Factors and Health Issues

• Investigating correlations between socioeconomic conditions and health issues (like respiratory diseases) can inform preventive strategies aimed at reducing incidence rates through educational programs on hygiene practices.

Types of Study Designs Based on Research Problems

• The design chosen by researchers should align with the type of problem being investigated:

• Descriptive Studies: Corresponding with descriptive problems.

• Association/Correlation Studies: Addressing association problems.

• Intervention Studies: Focused on comparing effects across different interventions or treatments.

Classification of Study Designs

• Research designs can be classified into three main groups:

• Observational Studies: Including descriptive designs, association studies, risk factor studies, and diagnostic method evaluations.

• Experimental Studies: Encompassing controlled and uncontrolled clinical trials.

• Explanatory Studies: Utilizing hypothetical-deductive methods for explanation-based inquiries. Understanding these classifications aids researchers in selecting appropriate methodologies based on their questions and objectives.

Observational vs Experimental Approaches

• Distinguishing between observational and experimental approaches is crucial:

• Observational studies involve recording events without manipulation.

• Experimental studies require active intervention by the researcher to influence outcomes (dependent variable). This distinction impacts how results are interpreted and applied within research contexts.

Understanding Research Designs in Quantitative Studies

Types of Research Designs

• Cross-sectional vs. Longitudinal Studies: A study is cross-sectional if variables are measured once per subject; longitudinal if measured multiple times, allowing for evolution assessment. The duration does not determine the type.

• Example of Longitudinal Study: Evaluating body temperature changes over 24 hours in hospitalized patients can be a longitudinal study conducted within a single day.

• Example of Cross-sectional Study: Assessing how many individuals from a neighborhood visit a specific health center is cross-sectional, even if data collection spans several days.

Retrospective vs. Prospective Studies

• Retrospective Design: Involves analyzing data collected before the research problem was defined, such as reviewing medical histories to address an issue.

• Prospective Design: Data is collected after defining the research problem, indicating planned and standardized measurements, like monitoring body temperature with pre-planned assessments.

Flow Diagrams in Research

• Purpose of Flow Diagrams: These graphical representations outline various stages of quantitative research designs, including population characteristics, measurement moments, and statistical inference results.

• Descriptive Study Example: For studying physical activity patterns in Chileans, flow diagrams should include demographic details and geographical information about the population sample being studied. Conclusions and statistical inference must also be documented clearly.

Measurement Considerations

• Single Measurement Implications: If only one measurement occurs (e.g., physical activity patterns), no additional arrows are needed on flow diagrams to indicate measurement frequency. This simplifies representation but limits depth of analysis regarding variable relationships.

• Multiple Measurements Example: For assessing clinical characteristics and evolution in children with severe bronchiolitis in ICUs, it’s crucial to specify how many measurements were taken and their timing (e.g., daily checks post-admission). This would require multiple arrows on flow diagrams to represent each measurement point accurately.

Observational Attitude in Research

• Researcher’s Role: In descriptive studies where variables are recorded without manipulation by researchers, the approach remains observational regardless of whether designs are cross-sectional or longitudinal based on data collection timing (retrospective or prospective).

• Association Studies Classification: Variables can be classified as independent or dependent based on their roles within association studies; for instance, accessibility may depend on residential area (independent variable). Understanding these relationships helps clarify study design implications for future analyses.

Understanding Epidemiological Study Designs

Types of Observational Studies

• Observational studies can be classified based on the number of measurements as either cross-sectional or longitudinal. They can also be categorized by the timing of problem formulation as prospective or retrospective.

• Risk factor studies are a type of epidemiological study that assess the likelihood of a population developing an event or disease due to prior exposure to specific risk factors.

Case-Control and Cohort Studies

• Two main approaches for evaluating associations are case-control studies and cohort studies. Case-control studies start with individuals who have the event (positive cases) and those who do not (negative controls).

• In a case-control study example, if investigating whether antipsychotic use is a risk factor for extrapyramidal syndrome in schizophrenia patients, researchers would compare past medication usage between both groups.

Analyzing Cohort Studies

• Cohort studies involve two groups of healthy subjects: one exposed to a risk factor and another not exposed. These groups are followed prospectively to see if they develop the event in question.

• For instance, examining nutritional status's relationship with hospital morbidity/mortality involves comparing hospitalized patients' nutritional states at admission and tracking outcomes over time.

Clinical Trials Overview

• Clinical trials assess therapeutic effectiveness and can be controlled or uncontrolled, depending on whether there is a control group receiving an alternative intervention.

• In clinical trials, participants may be divided into experimental groups receiving new treatments alongside standard care versus control groups receiving only standard treatment.

Diagnostic Test Evaluation

• Diagnostic test evaluation studies compare sensitivity, specificity, and predictive power against established gold standards using samples from healthy subjects and those with conditions.

• For example, assessing imaging techniques like MRI versus CT scans in stroke patients involves evaluating both methods across different patient samples.

Research Design Selection

• Identifying research problems leads to selecting appropriate designs tailored to address specific questions effectively.

• Research designs can vary based on objectives, investigator attitudes, measurement frequency, and data collection timing; careful planning is crucial for effective design choice.