Download Quantitative Research Designs & Methods: Experimental and Non-Experimental Approaches and more Exams Advanced Education in PDF only on Docsity!
Module ; Quantitative Designs
& Methods (Part I)
Overview of Quantitative Research Design - -design THEN do
- In quantitative studies, the research design has a significant impact on the validity of the results. oThe design determines the strategies that researchers must use to answer their questions and test their hypotheses.
- Several design decisions that researchers must consider when planning their study... -have to send a lot of time planning -do not have flexibility to change oImplications of these decisions = impact on the study's rigour quantitative design: Interventions - Key Question: Will there be an intervention introduced? If yes, how will researchers test its effects? Broad Quantitative Design Options
- Experimental (randomized control trial or RCT) -involves an experiment/intervention
- Quasi-experimental (controlled trial without randomization) -has an experiment
- Non-experimental (observational study) -no experiment
Comparisons - Key Question• What types of comparisons will be made to illuminate relationships?o Provide interpretive context design options for comparisons - Some Design Options o Within-participants design: Same people are compared at different times or under different conditions (e.g., pre-op and post-op) -same people compared to themselves o Between-participants design: Different people are compared (e.g., those getting versus not getting an intervention). -comparing two populations Cofounding Variables - How will confounding variables (extraneous factors that can impact study) be controlled? Which specific confounding variables will be controlled and how? Confounding variables are outside factors (besides dependant and independent variables) that can impact results eg age, pain etc. design options confounding variables - • Some Design Options o Randomization, crossover, homogeneity, matching Examples: Cofounding Variables - 1. comparing how infant birth weight (IV) influences maternal weight gain (DV) -also must take into consideration length of gestation (confounding variable)
- comparing how smoking (IV) can increase the risk for lung cancer (DV -also must look at workplace environment such as mining (counfounding factor)
Masking/Blinding - • From whom will critical information be withheld to avert bias? o E.g. Interventionists, data collectors, participants, etc. single blind: participants are do not know double blind: both participants and researchers dont know other quant key features - Location• Where will the study take place? o Setting choice (various) o Single-site vs multi-site Ø Timeframes How often will data be collected? When, relative to other events, will data be collected? o Cross-sectional (single point in time), longitudinal design (multiple points of data collection) Causality - Many (if not most) quantitative research questions are about causes and effects.
- Causes are seldom deterministic - they only increase the likelihood that an effect will occur. o For example, smoking is a cause of lung cancer, but not everyone who smokes gets lung cancer and not everyone with lung cancer has smoked
- An effect represents the difference between what happens when exposed to the cause and what would happen if they were not. Research Questions & Research Design - Different designs are appropriate for different questions. Experimental designs (RCTs) offer some of the strongest evidence of whether a cause (an intervention) results in an effect (a desired outcome).
o That's why they are high on evidence hierarchies for questions about causes and effects what is an experiment - A true experiment involves randomization and manipulation of the independent variable, while controlling for extraneous variables (confounders) and holding all other potential influencing factors (e.g. conditions of data collection) constant. Any other type of study is NOT an experiment! Experimental Design: Randomized Control Trials (RCTs) - 'Gold standard' for providing information about cause-effect relationships
- Intervention: The researcher does something to some participants — introduces an intervention (or treatment).
- Control: The researcher introduces controls, including the use of a control group.
- Randomization: The experimenter assigns participants to a control or experimental condition on a random basis. what is the purpose in a RTC - o The purpose is to make the groups equal with regard to all other factors EXCEPT receipt of the intervention Does this Exemplar Case Meet the Criteria for an Experimental Design? Why/Why Not?! You are investigating the effects of gentle massage compared to no massage on pain in nursing home residents. The residents are randomly assigned to the intervention and control group. A pre-test and post-test design was used to observe pain levels before and after the intervention. - YES! Because... üAn intervention is being introduced to some of the participants (manipulating the IV) to observe impact
measured in each group by an online, 30-min. proctored HESI© post-test developed for the study immediately after both teaching modalities. After-Only Experimental Design - -aka posttest-only design -no pre test, only tested at the end -there is a random assignemt, experimental group gets intervention and then both experimental and control group have a post -test Crossover Design - • Participants are exposed to 2+ conditions, in random order. • Participants serve as their own control. example of a crossover design - Example DiLibero and colleagues (2015) used a crossover design to test whether withholding or continuing enteral feedings during repositioning of patients affected their incidence of aspiration. The same patients were randomly assigned to different orderings of enteral feeding treatment. cross over design explained - -participants are randomized -they are divided into group A and B. In the first phase, group a gets intervention Y and group B gets intervention X. In phase 2 they switch interventions. They are evaluated after each intervention -both are getting same interventions in different orders Experimental & Control Conditions - • Researchers describe the intervention in formal protocols that stipulate exactly what the treatment is.
- Attention must be paid to intervention fidelity (or treatment fidelity), that is, whether the treatment as planned was actually delivered and received. Control Group Options: - a) No intervention is used; control group gets no treatment at all. b) "Usual care" or standard or normal procedures is used to treat patients. c) An alternative intervention is used (e.g., auditory vs. visual stimulation). d) A placebo or pseudo-intervention, presumed to have no therapeutic value, is used. e) Delayed treatment (e.g., wait-listed & exposed to the intervention at a later date)
Advantages to experimental designs - Most appropriate for testing cause- and-effect relationships Provide highest level of evidence for single studies asking cause/effect questions disadvantages to experimental design - -Not all research questions are amenable to experimental manipulation or randomization -Issues with feasibility/ethics (therefore cannot be conducted)-ethics can be difficult with placebo, sliding, safety etc -Challenging logistics in field settings -Expensive to conduct -hard to coordinate with large samples Quasi-Experimental Design - Involve testing an intervention BUT lack randomization (and sometimes, also lack a control group) -no randomization BUT still an intervention 3 types of quasi-experimental designs - 1. Non-equivalent control group design oThose getting the intervention are compared with a non-randomized comparison group. oIdentical to pretest-posttest experimental except for randomization
- Non-equivalent control group posttest only oIntervention & control groups but no baseline data
- One group pretest-posttest design oOne group is studied before and after the intervention.
-easier not to randomize Disadvantages
- Some generalizability - but more limited• Difficult to make clear cause-and-effect statements • Decreased validity of the findings Non-Experimental Studies - -no intervention or randomization, just observing
- If researchers do not intervene by controlling an independent variable, the study is non-experimental (observational).
- Not all independent variables ("causes") of interest to nurse researchers can be experimentally manipulated. oFor example: a neonate's birth weight cannot be manipulated; smoking cannot ethically be manipulated. Correlational Designs - - a type of non-experimental design Research that explores the interrelationships among the variables of interest, with no researcher intervention. o They are studying the effect of a cause they cannot manipulate Observing whether the variables co-vary (as one variable changes does a related change occur in the other variable?) very important point about correlation deigns - Correlational studies examine associations between variables. But the variables are not causational!
- Correlation does not prove causation! example of correlation is not causation - both ice cream sales and shark attacks increase when the water is hot but they are not caused by each other.
Descriptive Research - Not all research is cause-probing; some research is descriptive (e.g., ascertaining the prevalence of a health problem).
- The purpose of descriptive studies is to describe and document aspects of a situation. E.g. surveys! Survey Designs - Commonly used by health investigators. Provides information regarding the prevalence, distribution and possible associations of variables within a population.
- A variety of methods that involve asking questions, including questionnaire and interviews Advantages of Non- Experimental Research - Advantages:
- Efficient way to collect large amounts of data when intervention and/or randomization is not possible. Many nursing studies are seeking to address problems that do not lend themselves to intervention. Non-experimental work is often necessary BEFORE interventional studies can be justifiable. Disadvantage of Non- Experimental Research - Disadvantage: Does not yield persuasive evidence for causal inferences• This is not a problem when the aim is description, but correlational studies are often undertaken to discover inferences.• Can be susceptible to faulty interpretation. Timing: Cross-Sectional Design - Data is collected at a single point in time. Appropriate for describing the status of phenomenon or relationships among phenomenon at a fixed point... a "snap shot" Population sampled and exposure/disease status are measured at the same time. Many limitations, but commonly used (economical).
- Relatively inexpensive and fast to conduct, and often provides information used to design other studies. -not same strength of data since we are not following participants over time Longitudinal Design - • Longitudinal Design: Data are collected two or more times over an extended period.
