Nursing Research
glovesme
chapter_8.pptxQuantitative Research Design
Chapter Eight
Definition
Identified with the traditional scientific method
Gathers data objectively in an organized manner
Findings can be generalized to other situations/populations
Systematic plan
The most commonly used research design is
quantitative design
2
Types
Experimental designs
Treatment
Control group
Quasi experimental
Lacks randomization
May not have control group
Non experimental
Generates questions for experimental designs
3
Characteristics
Why – Purpose for the study
Where – Setting where the study is done
Who – Subjects to be studied
What – Type of data to be collected
When – Timing of the data collection
How – Design used for the study
These questions help look for a cause and effect in quantitative research
4
Experimental Design
Control
Most common and important characteristics
Manipulation
Independent variable
Randomization
Assignment of subjects to a group allowing for equal opportunity of selection
5
Variables
Dependent
Outcome of the study that is caused from the independent variable
Usually not manipulated
Independent
Treatment, intervention, or experiment
Usually manipulated
Extraneous
Need to be controlled
Examples – gender, age, ethnicity
6
Generalization
Ability to apply the results of a study to other settings and/or populations
Helps to strengthen the study
Aided by careful control of the study through the use of randomization, control, and manipulation
7
Descriptive Designs
Examines the characteristics of just one sample population
Used for theory development, practice problems, rationale for current practice or clinical decision making based on what others are doing
Types
Comparative
Time dimensional
Cross-sectional
Trends and events
Correlational
8
Comparative design
Looks at the differences in two or more groups
No manipulation of the independent variable
No control group used
Can be retrospective
9
Correlational Design
Most widely used descriptive design
Examines the relationships between two or more variables within a situation
Reason for the relationship is unknown
Do not conclude that only one variable causes another
Can be prospective or retrospective in design
Can be predictive in nature
10
Experimental Designs
Look for cause and effect
Must be a preceding cause and a relationship between the cause and outcome
Issues to consider
Can the variables be manipulated
Ethics
Feasibility
Hawthorne effect
Types
Pretest-posttest
Randomized Controlled Trials
11
Non Experimental Designs
Manipulation of variables is not possible
Randomization is not possible
Observational not interventional
Types
Secondary analysis: examines data from another study
Meta analysis: examines data from previous studies and compares them to several studies
12
Quasi Experimental Design
Most frequently used quantitative research design
Independent variable is manipulated
No randomization of subjects
No control group used
Not as strong design as experimental
Stronger design than descriptive
More practical
13
Control
Critical for results related to cause and effect in relation to the treatment/intervention
Randomization aids in controlling extraneous variables
14
Evidence-based Considerations
Increases comprehension of various designs
Concepts of randomization and control are essential
What is expected (relationships, effect, etc) determines design
Three essential components must be present for experimental design – control of variables, manipulation, and randomization
15
Summary Points
Identified with traditional scientific methods
Objective, systematic plan for gathering data
Design centers around why, where, who, what, when and how questions
Examines relationships for cause and effect
16
Summary Points - continued
Manipulation of the independent variable, control of extraneous variables, and randomization are essential
Comparative designs – no manipulation or control of the independent variable
Correlational design most commonly used descriptive design
17
Summary Points - continued
Experimental designs look for cause and effect
Issues – ethics, manipulation of variable, feasibility, and Hawthorne effect
Most classic experimental design – pretest-posttest design
Randomized Controlled Trials – true experimental design
18
Summary Points
Types of non experimental designs used in EBP – secondary analysis and meta analysis
Quasi experimental designs used most frequently when no randomization or control group possible
Non equivalent control groups design compares two groups without randomization
19
Summary Points - continued
Time series design – no randomization, no control group, data collected at various intervals.
Implications for EBP – must have working knowledge about quantitative designs.
20
