Research Design

Research

Systematic inquiries that helps us determine how things are and how they might be 

Research Characteristic: Systematic

Research follows structured, organized process with clear and defined steps 

Research Characteristic: Logical

The reasoning behind the research is sound and based on established principles 

Research Characteristic: Empirical 

Research is based on observable and measurable evidence

Research Characteristic: Reductive

Research simplifies complex fiascos by breaking them down into workable parts

Research Characteristic: Replicable

Research can be repeated and have similar results

Scientific Method Steps (5)

1. Define the problem

2. Formulate the Hypothesis

3. Gather data

4. Analyze results

5. Interpret findings

Research Type: Descriptive

observes and describes circumstances 

Research Type: Philosophical

explores meanings, ethics, and values

Research Type: Historical 

Investigates past events

Ethical Principles

1. Respect for persons

2. Concern for welfare

3. Justice

Key Principles of CBPR

• Collaboration

• Respect 

• Reciprocity 

• Empowerment 

• Transparency 

Informed Consent

Voluntarily agreeing after being full informed about outcomes and alternatives

Ethical Issues in P.A

• Vulnerable populations

• Use of Deception and Placebos

Deception

Misleading and untrusting that may be used in research to prevent bias

Placebos

Substance/treatment that has no therapeutic effect but is given to satisfy/psychologically benefit them 

Placebo effect

The positive change in a person’s health or symptoms that occurs post treatment with no active therapeutic ingredient

Review Type: Narrative

Descriptive summary of literature, often without a structured search

Review Type: scoping 

Maps key concepts, gaps, and types of evidence in a broad topic area

Review Type: Systematic 

A structured, comprehensive, synthesis of research studies on a certain question using a predefined tool

Review Type: Meta-Analysis

Statistical technique that mixes results from various studies to produce a pooled estimate

PICO(T)

• Population, Intervention, Comparison, Outcome, Time

• Helps researchers create focused, clear, and answerable research questions for interventional studies

PEO

• Population, Exposure, Outcome

• Useful for investigating relationships and trends without intervention

Advantages of AI

• Efficient

• Extensibility

• Accessibility

• Creativity

Importance of Measurement

allows researchers to quantify performance, track progress, and evaluate interventions

Variables

Anything measurable or observable that can change

Importance of Variables

Helps us measure the right things, the right way, to answer meaningful questions 

Independent Variable

Variable manipulated by the researcher

Dependent Variable

Outcome that is measured

Variable Type: Control

Keeps constant to avoid influencing the outcome

Variable Type: Extraneous

Uncontrolled variables that may affect the dependent

Variable Type: Moderator

Variable that influences the strength/direction of the relationship between the independent and dependent

Accuracy

How close a measurement is to the true value 

Validity

Dictates if the measurement tools are actually measuring what its supposed to

Validity Type: Face 

Test that measures what its supposed to at face value

Validity Type: Content

Test that covers all relevant aspects of the concept

Validity Type: Criterion

Test that correlates with a known standard

Validity Type: Concurrent

Comparison with a gold-standard test simultaneously

Validity Type: Predictive

Test is able to predict future performance 

Validity Type: Construct

measures the theoretical concept it claims to 

Reliability 

• Consistency

• Test is reliable if it produces the similar results when repeated under the same conditions

Reliability Type: Test-retest

Measures consistency over time

Reliability Type: Alternate Forms

Compares 2 different versions of the same test; useful for avoiding test-retest bias

Reliability Type: Internal Consistency

measures consistency within a test

Reliability Type: Inter-rater

measures agreement between different testers

Importance of Reliability

• Test must produce similar numbers

• Reliability is a prerequisite for Validity

Reliability-Validity Threats

• Poor test design

• Testing environment 

• Participant Factors

• Measurement tools 

• Social desirability bias 

• Tester variability 

• Scoring subjectivity

Precision

Minimizing error in measurements

Standard Error of Measurement

• Tells us how much a score might vary from a person’s true score 

• Smaller SEM = more precise measurement 

Quantitative Design

• Tests hypotheses and measures variables 

• Data Type: numbers 

• Analysis: statistical 

• Outcome: generalizable results

• Sample Size: larger representative

Qualitative Design

• Explores meanings and experiences

• Data Type: Textual

• Analysis: Content  

• Outcome: deep understanding of context

• Sample Size: Small and purposeful

Mixed Methods Design

• Combines strengths of both Qs

• Data Type: numbers and text

• Analysis: statistical and content  

• Outcome: comprehensive insight 

• Sample Size: Varies depending on design

Observational Study 

Involves observing study patterns, associations, and risk factors

Odds Ratios 

• A measure of association between exposure and outcome

• OR = 1 - No association

• OR > 1 - exposure increase the odds of outcome

• OR < 1 - Exposure decreases the odds of outcome 

Cross Sectional Study

• Survey of frequency of disease risk factors or other characteristics in a defined population at one point in time 

• Inexpensive, understands current status but can’t infer if behavior is the reason

Longitudinal Study

• Study that follows the same group of people over time to observe changes in exposures and outcomes 

• identifies patterns over time and reduces recall bias 

• Time consuming and expensive 

Case Control Study

• Study where characteristics of people with a condition are compared with a selected people without the condition

• Less expensive and time consuming 

• Measurements are taken after the disease has occurred 

Cohort Study

• Study of a group of people with a specific characteristic/condition who are followed over a period of time to find change 

• can determine exposure status before disease occurs 

• May be inefficient way to measure rare diseases that have long latency 

Quasi Experimental Study 

• Study in which an intervention is applied to one group and the outcome is compared from baseline to post treatment

• provides better evidence for the effectiveness

• Can’t account for non-treatment related outcome effects 

Prospective controlled Trials

• Pre and post with comparison group; 1 group gets treatment while other group is placed into non-randomized sequence 

• better understanding of the effect occurring 

• Still has no potential for placebo effects and confounders 

Experimental Research

Manipulates the independent variable to identify cause and effect 

Randomized controlled Trials

• Randomly placed into treatment and comparison

• Only study that can establish causation 

• Can be expensive and timely 

RCT: Conventional

One sample size placed into 2 groups to see outcome

RCT: Stratified

One sample size placed in 2 subsamples, then conducts study conventionally

RCT: Crossover 

Sample size gets placed into 2 groups testing 2 different things, then they rest, and then go to the next group

RCT: Cluster

Divides population into groups and then randomly selects entire clusters to participate

Sampling Bias

sample selected isn’t representative of the whole population

Attrition/Loss of data Bias

Occurs when data is missing or participants drop out of a study

Assessment bias

when the way outcomes are measured is influenced by factors unrelated to the variables studied 

Observer bias

Person assessing outcome is influenced by knowledge of the participant group expectations

Recall bias

Participants are asked to remember past events and their recall is influenced by their current conditions/beliefs

Response bias

Participants selectively report outcomes based on what they think is expected or socially acceptable 

Instrument bias

Tools/methods used to measure outcomes inconsistent

Deductive 

Starts with a theory then tests it with data (Theory, Hypotheses, Measure)

Inductive 

Starts with observations then builds theory (Interview then conclusion)

Qualitative purpose

• seeks to understand how people experience P.A

• Exploring attitudes, motivations, barriers, and cultural meanings 

Ethnographic research

Immersion into a culture or community to understand norms, values, and behaviors

Grounded Theory

generate theory from data rather than test existing ones

Phenomenological research 

focuses on the lived experiences of individuals

Interpretive research

emphasizes understanding behavior in natural settings

Critical Theory research

Focuses on power, inequality, and social justice

Trustworthiness Criteria

1. Credibility - confidence in findings

2. Transferability - applicability of findings

3. Dependability - consistency of findings

4. Confirmability - neutrality and objectivity of findings

Trustworthiness techniques

1. Triangulation - using various data sources 

2. Member checks - share findings 

3. Audit trail - maintain detailed records

4. Peer debriefing - discussing findings  

Mixed Methods Usage

• Complimentary - qual. data adds depth to quan. findings 

• Triangulation - confirms findings through various lenses 

• Development  - One method informs the other 

• Expansion - broadens the scope of inquiry 

Convergent Parallel

Quan and Qual data together, analyze separately, then merge results for interpretation

Explanatory Sequential

Use quan. data, then use qual. data to explain or elaborate on the findings

Exploratory Sequential 

Use qual. data to explore a scenario, then quan. to test/generalize findings

Multiphase

Combine multiple studies/phases over time, often for complex interventions/programs