Using Research and Statistics in Health Care
Fundamentals of Statistics and Terminology
Statistics Definition: Statistics is defined as the science of analyzing and learning from data.
Population: The entire group of individuals to be studied. In measurement terms, it represents the set of measurements corresponding to the entire group of individuals about which information is desired.
Individual: A person or object that is a member of the population being studied. Individuals are the objects described by a set of data and can include people, animals, or things.
Sample: A subset or part of the population from which data is actually collected. It is used to draw conclusions and infer characteristics about the broader population.
Census: A specific type of sample survey that attempts to include the entire population in the sample.
Variable: Any characteristic of an individual.
Categorization of Statistical Methods
Descriptive Statistics:
Involves numerical or graphic summaries of data.
Presentations and tools include charts, graphs, tables, and summary statistics (such as the mean and standard deviation).
Inferential Statistics:
Consists of statistical techniques that allow conclusions to be drawn about the relationships found among different variables.
Examples include the Chi-square test, t test, and Analysis of Variance (ANOVA).
Classification of Study Types
Descriptive Studies:
Purpose: Simply describe situations and events.
Research Questions: Ask descriptive questions (e.g., "What is the average length of stay in a hospital after being admitted for an asthma attack?").
Analysis Tool: Descriptive statistics are used to analyze data from these studies.
Explanatory Studies:
Purpose: Elucidate relationships among variables. They may or may not seek to establish causality.
Research Questions: Ask inferential questions (e.g., "Are women who are sedentary during the trimester of pregnancy more likely to have a C-section than women who exercise regularly during the trimester?").
Analysis Tool: Uses inferential statistics.
Prediction & Control Studies:
Purpose: Seek to determine which variables predict other variables and to determine causality.
Research Questions: Ask inferential questions (e.g., "Are people who receive an experimental medication less likely to have symptoms of the disease than people who receive the standard medication?").
Analysis Tool: Uses inferential statistics.
The Ten-Step Study Plan
Statement of the Problem and Its Significance:
Explains the purpose of the study, including the central research question, why it is important, and how it fits into the existing body of research.
Requires clear and concise formulation that explicitly states what the study seeks to accomplish (e.g., using phrases like "the purpose of this study is to…").
Theoretical/Conceptual Framework:
All studies require an underlying theoretical or conceptual framework to organize the analysis.
Draws from existing theoretical models or provides a logic model.
Uses the framework to organize research questions and hypotheses, providing appropriate citations.
Research Questions to Be Answered:
Stem directly from the statement of the purpose of the study.
Grounded in theory and current literature.
Explicitly state expected relationships and relate directly to the data that will be gathered by the researcher.
List of Hypotheses to Be Tested:
Definition: A hypothesis is a tentative prediction or description of the expected relationship between two or more variables.
Function: Translates research questions into statements that can be tested using inferential statistics.
Directional (One-Way) Hypotheses: State the specific expected direction of the relationship between two variables (e.g., "People who are immunized will be less likely to contract the flu than people who are not immunized.").
Nondirectional (Two-Way) Hypotheses: State that a relationship exists between variables without specifying the direction (e.g., "There will be a relationship between obesity and exercise level.").
Definitions of Key Terms and Variables:
Defines all terms that may be unclear to the reader and spells out all acronyms upon first use.
Defines all variables and specifies their roles:
Independent Variables: Variables that are manipulated and/or may affect the outcome of interest (typically including age, gender, and ethnicity).
Dependent Variables: The primary outcomes of interest (typically including health status, use of health services, and cost of care).
Covariates: Additional variables accounted for in the analysis.
Description of the Research Design:
Details how data will be or were collected.
Identifies the study design type (observational, quasi-experimental, or experimental).
Specifies data gathering methods (interviews, surveys, medical records, etc.).
For secondary data analysis (existing data), describes and cites the originating study.
Description of the Sample and How It Was Obtained:
Outlines the sampling method (random or nonrandom).
Specifies overall sample size and the size of each comparison group.
Details sociodemographics (age, gender, ethnicity, education, marital status, and other relevant variables).
Description of the Planned Statistical Analysis:
Explains data cleaning procedures to ensure data is error-free.
Identifies the planned descriptive and inferential statistics.
Specifies the statistical models to be built (e.g., linear regression, logistic regression, ANOVA, among others).
Statement of Assumptions, Limitations, and Delimitations:
Assumptions: Statements taken to be true without direct evidence (e.g., assuming participants report exercise frequency or body weight accurately).
Limitations: Weaknesses of the study that may limit result validity (e.g., small sample sizes, poor response rates, poor follow-up rates, or lack of random selection).
Delimitations: Boundaries to which the study was deliberately confined by the researcher (e.g., restricting inclusion to adults only, a specific age group, or women only), which directly limit generalizability.
Dissemination Plan:
Establishes how study results will be shared with external audiences.
Includes internal reports, conference presentations, trade journal publications, and peer-reviewed journal publications.
Practice Questions and Concept Verification
Question 1: Explanatory studies simply seek to describe situations and events. (True or False)
Answer: False
Rationale: Explanatory studies seek to elucidate relationships among variables rather than merely describing situations and events.
Question 2: Descriptive studies do not need a framework. (True or False)
Answer: False
Rationale: All studies need to be organized using a theoretical or conceptual framework.
Question 3: Which of the following is an example of an assumption of a study?
A. The self-reported body weight was accurate.
B. Only women ages to were included in the study.
C. The response rate to the survey was only .
D. All of these are assumptions.
Answer: A. The self-reported body weight was accurate.
Rationale: Expecting subjects to correctly report their body weight is an assumption. Option B is a delimitation (a deliberate boundary) and option C is a limitation (a weakness impacting validity).