UNDERSTANDING QUANTITATIVERESEARCH DESIGN Dr. Anne Snowdon Professor Lisa High University of Windsor QUANTITATIVE RESEARCH - based on the philosophy of “logical empiricism” or “positivism” - Application of logical principles and reasoning - Researcher adopts a distant and non-interactive posture = to prevent bias DEFINE – QUANTITATIVE RESEARCH It is a formal, objective, systematic process in which numerical data is utilized to obtain information about the world It is a method used to describe variables, examine relationships among variables, and determine causeand-effect interaction between variables Quantitative Research Design Involves decisions: Will there be an intervention? What types of comparison will be made? What procedures will be used to control extraneous variables? When and how many times will data be collected from study participants? In what setting will the study occur? Types of Quantitative Designs Designs & Major Features: Control over the “independent variable” (a) Experimental (b) Quasi-experimental (a) Non-experimental Designs & Major Features Designs & Major Features: Type of “group comparisons” (a) Between Subjects (b) Within Subjects Designs & Major Features Designs & Major Features: Number of “data collection points” (a) Cross-sectional (b) Longitudinal Designs & Major Features Designs & Major Features: Occurrence of “independent and dependent variable” (a) Retrospective (b) Prospective (c) Descriptive Designs & Major Features Designs & Major Features: “Setting” (a) Naturalistic (b) Laboratory Quantitative Designs Discussion EXPERIMENTAL DESIGN Characteristics: (1) Manipulation – treatment/intervention (2) Control – control group/experimental group (3) Randomization – random assignment – no systematic bias Quantitative Designs Discussion Types of experimental designs: (a) (b) (c) (d) After-only or Post-test only Factorial Design Repeated measures design Clinical trials Experimental Design Advantages Most powerful Controlling properties Greater corroboration – independent variable affects the dependent variable Disadvantages Number of variables are not amenable to manipulation Many variables could technically be manipulated but not ethically Not feasible due to environmental aspects Hawthorne Effect QUASI-EXPERIMENTAL Types of quasi-experimental designs: Two most common are: (a) Nonequivalent Control-group before-after design (Pre-test Post-test) (b) Time Series Designs (one-group before-after design) QUASI-EXPERIMENTAL DESIGN Advantages Disadvantages Practical – not feasible to conduct true experiments Lack randomization or control group features Introduces some control – full experimental rigor is not possible Cause-and-effect inferences cannot be made easily – may alternative explanations for results “Rival hypotheses”- one plausible rival explanaton NON-EXPERIMENTAL RESEARCH Types of non-experimental designs: (a) Ex Post Facto (Correlational research) (b) Retrospective studies (ex post facto investigations) (c) Prospective studies (case-control design) (d) Descriptive (descriptive correlation study) NON-EXPERIMENTAL RESEARCH Advantages Disadvantages Nursing – problems not amenable to experimentation Inability to reveal causal relationships conclusively Correlational efficient & effective means of large data collection Self-selection Realism is high vs. artificial Quantitative Research Designs and the “Time Dimension Time Series Studies four situations: 1. Time-related processes 2. Time sequenced phenomena 3. Comparative purposes 4. Enhancement of research control CROSS-SECTIONAL STUDIES Cross-sectional Studies - data collection at one point in time - Appropriate for describing the status of a phenomena or relationships among a phenomena at a fixed point - Retrospective studies are usually cross-sectional - Weaker than longitudinal studies - Main advantage – economical and easy to manage LONGITUDINAL STUDIES Longitudinal Studies - data collected over an extended period of time - Main value – to demonstrate changes over time – temporal sequencing = causality - 3 types: (1) trend studies (2) panel studies (3) follow-up studies Attrition – may cause problems SPECIFIC TYPES OF QUANTITATIVE RESEARCH SURVEY -self reports EVAULATIONS - find out how well a program, treatment, or policy works - Types of evaluations: (1) process analysis (2) outcome analysis (3) impact analysis (4) cost-benefit analysis Outcomes Research - Nursing and allied health disciplines are involved - Overlaps with evaluation research - Purpose to document the effectiveness and efficiency of today’s health care system – to justify care practices, to improve patient outcomes and costs - A very global assessment of health care - Derived from quality assessment and quality assurance RESEARCH CONTROL TECHNIQUES - - main purpose of quantitative research – maximize control over the research situation Confidence that the conditions of the environment are not influencing the data maintain a true relationship b/w independent and dependent variables Control external and intrinsic factors CONSTANCY OF CONDITIONS: controlling “external factors” (1) (2) (3) minimize situational contaminants – laboratory vs. natural setting? time control – time of day or year data collected Communications with study participants – formal research protocols RESEARCH CONTROL TECHNIQUES Methods of controlling “Intrinsic Factors”: (a) Randomization (b) Homogeneity (c) Matching (d) Statistical control INTERNAL & EXTERNAL VALIDITY INTERNAL VALIDITY Define – the extent to which it is possible to make an “inference” that the independent variable is truly influencing the dependent variable - high degree of validity – experimental vs. quasi- experimental/correlation why? INTERNAL & EXTERNAL VALIDITY Threats to Internal Validity: (1) History (2) Selection (3) Maturation (4) Mortality INTERNAL & EXTERNAL VALIDITY EXTERNAL VALIDITY Define – the “generalizability” of the research findings to other settings or samples - adequacy of the sampling design - environmental or the research situation influences a study’s external validity Critiquing Quantitative Research Designs See page 198 – Box 8-1 Suggest to refer to for you term paper MAKING CONNECTIONS IDENTIFY THE “INDEPENDENT & DEPENDENT” VARIABLE (a) (b) (c) (d) (e) (f) Nurses’ attitudes toward mental retardation vary according to their clinical speciality area. The type of nursing care primary versus team is unrelated to the frequency of turning patients. The incidence of decubitus ulcers is related to the frequency of turning patients. Baccalaureate and associate degree nurses differ in use of touch as a therapeutic device with patients. First time blood donors experience greater stress during the donation than donors who have given blood previously. Surgical patients who give high ratings to the informativeness of nursing communications experience less level preoperative stress than do patients who give low ratings.