Non-Experimental Designs Notes

Non-Experimental Designs

Introduction

• Healthcare workers aim to:
  • Predict variable effects.
  • Examine variable causality to select optimal patient interventions.
• Nursing often studies cause-and-effect relationships but manipulating the independent variable (e.g., stress, pain) may be unethical or impossible.
• Non-experimental designs allow researchers to examine variables in a safe manner.

Non-Experimental Design Definition

• Used when researchers want to:
  • Construct a picture of a phenomenon at one point in time or over a period of time.
  • Explore people, places, events, or situations as they naturally occur.
  • Test relationships and differences among variables.
• Example: Studying pain.
  • Experimental research might intensify pain.
  • Non-experimental design examines factors that naturally contribute to pain (e.g., coughing).

Key Features

• The independent variable is not manipulated; it occurs naturally.
• Researchers explore relationships and differences between variables.
• Requires a clear research problem or hypothesis.
• Controls are implemented to ensure high-quality results.

Types of Non-Experimental Designs

• Broadly divided into:
  • Survey studies
  • Descriptive
  • Exploratory
  • Comparative
  • Relationship/Difference studies
  • Correlational
  • Developmental
    • Cross-sectional
    • Longitudinal or prospective
    • Retrospective or ex post facto

Survey Designs

• Used when little is known about the variables.
• Collect detailed descriptions of variables to explore/assess current practice conditions.
• Used to plan improvements in health practices.
• Survey: A data collection tool using questionnaires and interviews.
• Example: Exploring midwives' experiences in rural Northern Ontario.

Survey Designs - Keywords and Advantages/Disadvantages

• Keywords: exploratory, descriptive, comparative, or survey.
• Advantages:
  • Gathers detailed information from a large population economically.
  • Data is accurate if the sample represents the population.
• Disadvantages:
  • Information may be superficial.
  • Requires expertise in questionnaires, interviewing, and data analysis.
  • Can be time-consuming.

Relationship/Difference Studies

• Correlational studies:
  • Investigate relationships between two or more variables.
  • Examine the type and strength of relationships.
  • Do not examine cause and effect.
  • Example: Examining relationships between fatigue, resilience, and job satisfaction in nurses.
  • Often descriptive.
  • Can form a base for experimental or quasi-experimental studies.

Developmental Studies

• Examine relationships, differences, and status of variables or phenomena at a point in time.
• Examine how variables/phenomena change over time.
• Research designs:
  • Cross-sectional
  • Longitudinal or prospective
  • Retrospective or ex post facto

Cross-Sectional Design

• Examines groups of subjects in various developmental stages simultaneously.
• Aims to infer trends over time.
• Data is collected once from participants.
• Example: Attitudes of Canadian palliative care nurses regarding medical assistance in dying.
• Advantages & Disadvantages:
  • Examines attitudes at one moment in time.
  • Requires a good sample to reflect population beliefs.
  • Helps develop a foundation for examining variables

Longitudinal Design

• Panel designs examine the same subjects over an extended period.
• Involves multiple data collections from the same sample over months or years.

Longitudinal Studies - Advantages & Disadvantages

• Advantages & Disadvantages:
  • Effective for assessing developmental or long-term changes/implications.
  • Threats to internal validity: testing and attrition.
  • Costly in time and effort.
  • Presence of confounding variables.
  • Hawthorne effect can impact results.

Longitudinal Study Example

• Nurses' Health Study:
  • Determines relationships of hormonal, reproductive, dietary, lifestyle, biochemical, and genetic factors with coronary heart disease risk in female registered nurses.
  • Study type: observational, cohort.
  • Subjects: 121,700.
  • Time: prospective.
  • Start: August 1980; End: March 2024.

Retrospective or Ex Post Facto Design

• Determines relationships between past events.
• Works backward: the researcher knows the outcome and tries to determine the antecedents.
• Example: Is 12-hour shifts associated with increased sick absences for nurses?
  • Analyzed all RN shifts over 3 years and looked at sick call-ins.
  • Found 86% of staff had at least 1 episode of sickness, most commonly two days long

Retrospective or Ex Post Facto - Advantages & Disadvantages

• Advantages & Disadvantages:
  • Offers more control than correlational studies.
  • Difficult to determine causal links.
  • Alternative hypotheses may explain relationships.
  • Finding similar naturally occurring groups is challenging.

Other Types of Quantitative Research

• Methodological research:
  • Focuses on developing the validity and reliability of research instruments.
  • Does the Anxiety scale truly measure anxiety?
  • Is the instrument reliable? Does it consistently measure what the concept or construct is?
  • Psychometric research focuses on theories and techniques involved in measuring psychological constructs, especially measurement tools.

Systematic Review & Meta-Analysis

• Systematic review:
  • A structured, comprehensive synthesis of quantitative studies on a specific topic.
  • Determines the best research evidence for expert clinical use to promote evidence-based and evidenced-informed practice.
• Meta-Analysis:
  • Statistical pooling of results from several studies into a single quantitative analysis.
  • Highest level of analysis of evidence on intervention efficacy.

Integrative Review & Secondary Analysis

• Integrative review:
  • A critical review of a research area without statistical analysis or theory synthesis.
  • Example: methodology review, scoping literature review.
• Secondary analysis:
  • Reanalyzes data from a previous study for a secondary purpose.
  • Example: A survey on sexual health and depression; secondary analysis examines age, education, race, and employment related to situational depression.

Epidemiological Studies

• Explore factors affecting health and illness in a population related to the environment.
• Examine rates and characteristics of disease/illness.
• Do exposures to certain environments affect disease rates?
• Investigates the distribution, determinants, and dynamics of health and disease.
• Often prevalence or incidence-focused.

Research Design Comparison

Review

Assess the nature of the problem being studied

Little is known about variables

• Survey
• Descriptive
• Exploratory

Test relationship between variables

• Correlational

Compare differences between variables

*Comparative

Need a measurement tool

• Methodological

Past View

• Ex post facto
• Retrospective

Present View

• Survey
• Correlational
• Cross-sectional

Present Future View

• Longitudinal
• Prospective