Critical Thinking: In-Class Flashcards
Why is it important to use the scientific method?
It provides a systematic way to uncover truths!
There are cognitive biases and heuristics in thinking and reasoning…
- Hindsight bias: Looking back, we could have predicted…
- Overconfidence: Thinking we know more than we do.
- Perceiving patterns in random events: Built-in eagerness to make sense of the world.
- Availability heuristic: Tendency to think of the most readily available, vivid reason.
- Confirmation bias: Seeking information that confirms our current beliefs –> See relationships when there are none.
What are cognitive biases?
Systematic ways of thinking that may cause errors.
What are heuristics?
Mental shortcuts in thinking and reasoning.
What defines a science?
- Empiricism: Using observations
- Verifiability and replicability: Availability to the public
- Falsifiability
What type of process is the scientific method?
An ONGOING process!
Develop theories consistent with most or all currently available data
–>
Think of interesting questions relevant to the theory
–>
Formulate hypotheses, specific predictions that stem from the theory
–>
Come up with a design to test the hypothesis and establish operational definitions
–>
Gather, analyze, interpret the data
–>
Support or reject the hypotheses and the theory;
Further refine, alter, expand, or abandon the theory
On which two dimensions is the quality of data measured?
RELIABILITY
- Consistency in measurement: Amount of random error
- Generalization across test times and participants
VALIDITY
- Measuring what you intend to measure
- Generalizing across and beyond contexts, cultures, and tests
What are the different types of research design?
DESCRIPTIVE
- Naturalistic observation
- Case study
- Survey/self-report
CORRELATIONAL
EXPERIMENTAL
What are the steps within the scientific method?
- Form a theory, define a question, generate a hypothesis
- Operationalize variables; collect, analyze, interpret data
- Evaluate hypothesis, evaluate the theory.
- Revise and repeat!
Research design: Naturalistic observation
Observing and recording behaviour in a natural environment.
Pros:
- Natural setting = High validity: Matches the real-life situation
- Complex data = New ideas
Cons:
- Little control over the environment, limited measurement = Low reliability
- Difficult to determine cause/relation; description of behaviour only
- Danger of experimenter reactivity: If subjects are aware that you are observing, they may behave differently
- Danger of demand characteristics: Subjects believe you have a reason to observe, and try to meet your demands
- Danger of experimenter bias: Lack of control and lack of protocol
- Low validity
- Difficult to replicate
Research design: Case study
An in-depth qualitative and/or quantitative investigation on a single subject.
Pros:
- Unique opportunity
- Detailed and thorough
Cons:
- Small, non-random sample = Poor representativeness of a population = Low reliability and low external validity
- Difficult to replicate
- Danger of experimenter bias: You ask them questions seeking a specific response = Low validity
Research design: Survey/Self-Report
Interview or questionnaire with a large group of individuals with the goal of determining trends in a larger population
Pro: Easy collection of an abundance of data
Cons:
- Sampling bias: People are chosen in a systematic way, given unequal chances
- Availability heuristic
- Low reliability, iterates the importance of random sampling, using a sufficiently large sample
- Wording effects and question framing = Low validity (different interpretations)
Research design: Relational research
Observing and measuring two or more variables, examining their relationships
Pros:
-Reliable, replicable data
-Some variables cannot be manipulated, how are they taken into account?
For example, anything inherent about the individual; variables unethical to manipulate
Cons:
- Unknown direction of causation
- Third variable problem = We cannot draw conclusions about causation
How do we present relational research?
Using scatterplots, correlation coefficients
What matters? The correlation strength and direction
Research design: Experimental research
Manipulating one or more variables (independent variable), while observing/measuring the other variables (dependent variable). They use an experimental group and a control group.
Pros:
- Reliable, replicable data
- Can infer causation
Cons:
- Low ecological validity
- Some variables are impossible, difficult, unethical to manipulate
- Possible CONFOUNDING variables, which threaten internal validity
- Possible experimenter bias or demand characteristics: Use DOUBLE-BLIND procedures
- Sample representativeness = Threat to external validity
What is the key feature of experimental research?
Random assignment!
