Quantitative vs Qualitative Flashcards
Quantitative
- Theory is developed prior to research conduction
- Hypothesis testing to support theory
- Data = numbers
- Can manipulate data to show cause/effect information precisely
- Focus on controlling extraneous factors
Qualitative
- No hypotheses, see what happens
- Research is exploratory
- Data = words
- Deriven from an interview which is analyzed for emotions, expression and perceptions
- Combination of questionaire or observation with interview
- Prescence of researcher is manipulation of data; Do not manipulate the data with this type
- Focus on embracing extraneous factors
Quantitative research uses ____ reasoning
deductive
hypothesis testing
Qualitative research uses ____ reasoning
inductive
theory generating
Is which method of research, quantitative or qualitative, is the impact of the researcher very important?
- Qualitative!
- Researcher has direct impacts
In quantitative research we expect the majority of research to represent a ____
- Bell curve
- May have two subgroups within a bell curve.
Prospective research design
Research question developed before data collection
Data collection before development of condition or injury
Only way to determine etiology of injury
Retrospective Research Design
- Research question developed after data collection
- Data collection after development of condition or injury
- Unable to discern etiology
How can we compare data points?
- Time series or Discrete
- Time Series: Changes of one data point of a period of time)
- Discrete: One data point in one moment of time (single point)
We can simplify ____ to get ____ but not vice versa
- Time series
- Discrete
What are the 4 Characteristics of Quantitative Research?
- Manipulation
- Randomization: Relevant to both sampling and assignment
- Randomization: Random assignment to Tx/Group
- Control
Manipulation
- Characteristic of true experimental design; Manipulate controls
- Impose a set of predetermined experimental conditions; known as independent variables
- Independent variable is the presumed cause of a given effect
The more ____ validity, the less ____ validity
- internal
- external
Controlling more variables increases ____
- Internal Validty
What is an independent variable?
What types of independent variables are there?
- Independent variable: What you change that is different from the control.
- Two Types:
1. Active variables (Manipulated by the experimenter; You change it as researcher; Assignment to group)
2. Attribute variables (Naturally occuring groups (not manipulated; Age, gender, time, occupation)
What is a dependent variable?
- What is measured to determine the effect or outcome
Which is the dependent and independent variable?
Blood pressure and Type of Tx
Independent: Type of Tx
Dependent: Blood Pressure
Multiple IV are called ____
- Factorial designs
- These factors have levels
Example:
* Resistance Training (IV) Levels: Plyometric, Isometric, Eccentric
* Gender (IV): Males, Females
* Strength: 1 RM (DV)
Give an example of a 2 x 2
BFR - Balance and Strength
IV: BFR/No BFR, Pre/Post
DV: Nm (force), Sway (AP, ML)
Randomization: Sample and Assignment
- Sampling: How you choose your participants from population of interest (influences external validity - can this be generalized to the public?)
- Assignment: How will groups be split up of Tx vs non Tx (influences internal validity - similar in beginning so IV is more likely to be the cause of the difference in the DV at the end of the study)
Randomization: Random Assignment to Tx/Group
- Sample not necessarily randomly selected but randomly assigned
- Random number generator is a good option to limit bias
- Randomization of small samples may result in groups differing on important characteristics
- Groups likely have different size or same measures
When you can’t control other factors you may perform ____
- Randomization: matched assignment
- Paired baseon on some criteria to artifically contol the influence on the DV
- Common: BMI, age and gender
- Not a repeated measure design. Harder to find differences. Helpful for information on small groups of people
EX:
* A randomized matched design was used to randomly assign patients to either an eccentric exercise group or a standard rehabilitation group. Participants were matched by graft type, sex, and age.
* Graft type, sex, and age are controlled to be similar/same and then one person is placed in one group (eccentric exercise group) and another is places in a different group (standard rehabilitation group)
Control Groups
- Most effective deisng element to eliminate confounding variables
- Improves internal validity
What is the interaction effect?
When one group is plotted, and other group is plotted; means one of the factors is going to have a different response to treatment. Dependent variable is interacting as a result of one of our independent variables.
The control group gets ____ and the experimental group gets ____
- Standard Care
- Standard Care PLUS…
Types of Controls (5)
- Systematic manipulation of the IV (When, where, what type, by whom)
- Selection and assignment of subjects (inclusion/exclusion criteria)
- Confounding Variables (Factors other than IV that may effect the DV)
- Measurement error (Reliable and Valid Measures)
- Knowledge of the researcher or participant (Masking/Blinding)
Give an example of Masking/Blinding
Sugar Pill vs Real Med
Too much control lead to…
- Results only applying to limited conditions and subjects
- High Internal Validity, Low External Validity
Too little control leads too…
- Increased likelihood that an alternative explanation exists for the observed result
- High External Validity, Low Internal Validity
Types of Research Designs (3)
- Experimental (best for cause/effect; internal validity)
- Quasi-experimental (lack random assignment or comtrol/comparison group or both)
- Non-experimental (Data naturally exits; Worst for cause/effect, no random assingment or manipulation, may have a control/comparison group)
What are the 3 characteristics of experiemental design?
- Manipulation
- Control Group
- Random Assgnment
Best for Cause/Effect!
Quasi-experimental Designs
- Randomization is missing
- Consider affect of unknown group differences on changes in DV after Manipulation (Limits Internal Validity)
Quasi-Experimental Design:
Before-After with Multiple Follow-Up Tests
- No control
- Comprare within an individual
- Repeated measure of the dependent variable
- Care about the follow up tests more than pre test
Descriptive Designs
- No manipulation of the IV
- Can’t do cause/effect relationships
- Observaitonal; No change involved
Types of descriptive designs
- Correlational Studies
- Cross Sectional
- Survey Research
- Case Study
- Methodological Research
Epidemiology is a ____ design
non-experiemental
How is epidemiology analyzed?
- Prevalence
- Incidence
Prevalence
% of population with a certain condition at a given point in time
Incidence
Rate that a new cases are developing during a time period