GO

Quantitative Research Design

Plan for answering the research question

  • Quantitative Research Design provides a plan for answering the research question.

  • Includes plan for:

    • study setting

    • participant selection

    • study procedures

    • variable measurement

    • data collection

    • data management

    • data analysis

Key considerations in quantitative studies

  • Intervention: whether an active manipulation is applied.

  • Comparisons: what the control or comparison condition is.

  • Potential confounding variables: factors that may distort the observed effect.

  • Controlling study context: methods to minimize environmental or procedural variability.

  • Controlling participant factors:

    • randomization -randomly put participants in a group.

    • homogeneity - make sample similar

    • matching- match on a variable like gender, if theres a women in group a there’s on in group bThis technique ensures that both groups are comparable, thereby minimizing the effects of that variable on the study's outcomes.

    • statistical control

  • Blinding: masking participants, researchers, or assessors to group allocation when possible. In Drug Trails the Pharmacist may be unblinded in case a participant has a problem

  • Data collection times:

    • cross-sectional (Collect data all at one time) vs longitudinal designs (over a LONG period of time)

  • Relative timing:

    • prospective (forward) vs retrospective (back…example chart reviews)

  • Location: setting of data collection (labs, clinics, community, etc.).

Study designs (overview and evidence levels)

  • Systematic Review

    • Level I: Highest level of evidence

    • Level II for Therapy questions

    • Single RCT: Level II for some Etiology questions

    • Single Non-Randomized Trial (Quasi-Experiment)

    • Single Prospective Cohort Study

    • Level II for Prognosis questions

    • Level II for some Diagnosis questions

    • Single Case-Control Study

    • Single Cross-Sectional Study (e.g., a Survey)

    • Single In-Depth Qualitative Study

    • Level II for Descriptive quantitative questions

    • Level II for Meaning/Process questions

    • Expert Opinion, Case Reports, etc.

Classifications of study designs

  • Experimental

  • Quasi-experimental

  • Nonexperimental

Experimental designs (basic structure)

  • Key elements:

    • Intervention

    • Control

    • Randomization

    • Experimental group

    • Control group

  • Notation used in diagrams (conceptual):

    • Randomization = R, Intervention = X, Observation/Outcome = O

    • Examples of common designs include post-test only, pre-test–post-test, and crossover designs.

Experimental designs: common formats and notations

  • Post-test only design

    • Experimental group: R X O

    • Control group: R O

  • Pre-test – post-test design

    • Experimental group: R O X O

    • Control group: R O O

  • Crossover design

    • Structure involves sequences with periods where participants receive multiple treatments in different orders (randomized order) and cross-over between conditions.

    • Notation shown in slides: R O O O X O R O X O O O (illustrative representation of sequence changes across periods)

  • Advantages of experimental designs

    • Testing cause-and-effect relationships

  • Disadvantages of experimental designs

    • Balancing control with practical significance

    • Generalizability concerns

    • The need to randomize participants

Quasi-experimental designs (non-randomized manipulation)

  • Common structure: Intervention vs. Control without randomization

  • Notation reflects absence of randomization and the non-equivalent groups

Quasi-experimental designs: common types

  • Nonequivalent control group post-test only design

    • Experimental group: X O

    • Control group: O

  • Nonequivalent control group pre-test – post-test design

    • Experimental group: O X O O

    • Control group: O O O O

  • One group pre-test – post-test design

    • Group: O X O

  • Time series design

    • Multiple observations before and after an intervention: O O O X O O O

Quasi-experimental designs: advantages and disadvantages

  • Advantages

    • Practical

    • More acceptable to participants

    • Can be used when randomization is unethical

  • Disadvantages

    • More difficult to make causal inferences

Nonexperimental designs

  • Structure: Interventions vs. controls are not manipulated or randomized; observational in nature

  • Common types:

    • Correlation studies: examine relationships between variables that are not manipulated

    • Note: Correlation does NOT prove causation

    • Cohort designs

    • Case-control designs

    • Descriptive studies: observe, describe, and document a phenomenon

Nonexperimental designs: practical considerations

  • Advantages

    • Very practical

    • Efficient for collecting large amounts of data

  • Disadvantages

    • Cannot make causal inferences

    • Self-selection can bias results

Critiquing a quantitative study design (general approach)

  • Key questions to ask:

    • Does the design match the research question/hypothesis?

    • What are the strengths of the design?

    • What are the limitations of the design?

Validity considerations in critique

  • Types of validity to consider:

    • Statistical conclusion validity

    • External validity

    • Construct validity

    • Internal validity

Internal validity: threats to credibility of causal inferences

  • Common threats (examples):

    • Temporal ambiguity

    • Selection

    • History

    • Maturation

    • Mortality/Attrition

    • (Other common threats include instrumentation, diffusion, testing, regression to the mean, etc., though not listed explicitly in slides.)