Other terms Flashcards
Reproducibility
Whether the same result can be produced by using a different coding/analysis (SAME research method) and conducted by a different researcher (essentially two different tests measuring the same outcome)
Repeatability/Replicability
Whether the same result can be produced by replicating the same experiment but by different researcher
What is true causation?
True causation can only be established when necessity and sufficiency criteria are satisfied
Necessary
Y is necessary, but not sufficient to cause X
Sufficiency
Y alone is both necessary and sufficient to cause X
Population
A target group that contains a common characteristic that you want to research about
Sample
A small portion taken from a target group to conduct study on
Sample statistics
Characteristics about the sample
-> can be used to infer population parameter
Population parameter
Characteristics about the population
Differences between quantitative vs. qualitative research
Quantitative:
- use statistics
- usually through survey, observation, experiment
- use to test or confirm a hypothesis
- usually need lots of participants
Qualitative:
- use words and meaning
- usually through interview
- use to understand a theory
- usually don’t need lots of participants
Construct
refers to a concept or characteristic that can’t be directly observed but can be measured by observing other indicators that are associated with it.
(e.g. description or diagnosis of depression)
Theory
a collection of statements that together attempt to describe and explain a set of observed phenomena
-> usually provide a model to explain
-> makes general prediction
Hypothesis
a prediction made based on the theory of the observed phenomenon
criteria:
- falsifiable
- testable
- precisely stated
- rational
- parsimonious
Variables
any characteristic that can assume multiple values (i.e. can vary)
those that researchers manipulate or measure in experiment
need to be operationalized to measure
Confounding variable
a type of extraneous variable that can directly affect the DV instead of IV
(disproportionately affect one level of the IV more than the other
can result in us measuring:
- an effect of the IV on the DV when it is not present
- no effect of the IV on the DV when it is present
Extraneous variable
a type of variable that if failed to be controlled by experiment, can affect the DV measurement
Replication crisis
Methodological crisis in which the
results of studies were not
reproducible when tested again.
Benefits of open science
- Accumulation of knowledge
- Increased citation of work
- More media coverage
- Support meta-research practice
Open Science
The act of sharing research to the public (unrestricted public access)
APA open science conduct code
require researchers to be willing and able to make data available for 5 years after publication
What to share:
- Data
- Protocols
- Code for experiments and analyses
Benefit:
- Verification of methods
- Analytics Reproducibility
Pre-registration to publish
Researchers encouraged to submit plans for the specific research questions that they wish to address, and the analyses they will conduct prior
to data collection
What are good principles to conduct reproducible analysis?
- Provide clear annotations of what documents are
- Store original data files separately
- Record all steps of data processing
- Use open source software where possible
Steps to Induction
- Evidence from observations
- Conclusion is drawn
- Generate theories
-> can go on to test this
Hypothetical-deductive methods
- Theory taken from induction (observation and intuition)
- Form hypothesis
- Using empirical tests
- Gain results
-> supported: uphold hypothesis
-> unsupported: refined or abandon hypothesis
Bayesianism
- answer to induction and falsification
- involves calculating and updating probabilities as new information becomes available to make the best possible predictions.
Falsifiability
- answer to the problem with induction
- for a theory to be scientific, it has to be falsifiable (aka can be proven false by scientific methods)
What are the 5 characteristics of a good scientist?
- Uncertain
- Skeptical
- Open-minded
- Cautious
- Ethical
Relational/correlational research
- Study the correlation between 2 variables.
- Ideal for gathering data quickly from natural settings -> can generalize your findings to real-life situations
- Cannot test for causal-links
- Use when conducting experiment are not possible (e.g. unethical)
Descriptive research
- research aim to identify characteristics, frequencies, trends, and categories.
- does not control or manipulate any variable
- common: observation, survey, case studies
Construct vs. Variable
Construct is defined by theoretical definition (hard to observe or test)
Variable is defined by operational definition (easy to test)
Sample size
- Size matters!
- Sampling error can result if your sample is not large enough
- Trade off between size and time/cost
- Factors in deciding on sample size: design, response rate, heterogeneity of population
Random error
chance fluctuations in our measurement -> obscure the result (make less precise/accurate)
Systematic error
a bias is present and influencing our
measurement in a systematic way
-> bias the result